Chatbots in the Healthcare Industry

Version	Summary	Created by	Modification	Content Size	Created at	Operation
1		Eva Maia	--	2217	2023-10-23 17:39:30	\|
2	Reference format revised.	Lindsay Dong	-1 word(s)	2216	2023-10-24 02:18:14	\|

This entry is adapted from the peer-reviewed paper 10.3390/healthcare11182532

Chatbots have become increasingly popular in the healthcare industry. In the area of preventive care, chatbots can provide personalized and timely solutions that aid individuals in maintaining their well-being and forestalling the development of chronic conditions.

chatbot preventive care COVID-19 dementia artificial intelligence

1. Introduction

Chatbots are natural language processing systems that act as a virtual conversational agents mimicking human interactions. In recent years, the interest in chatbots is increasing, and they are being adopted by companies in several sectors, such as real estate, travel, education, healthcare, and finance ^[1]. More generally, chatbots such as Amazon Alexa, Google Assistant, Apple Siri and more recently ChatGPT have become popular among people, since they can support them to complete their daily routine tasks efficiently.

The growing interest in this kind of technology, allied with the need to support patients at home, triggered the development of chatbots in the healthcare sector. Using either rule-based or artificial intelligence (AI) methods, health chatbots can help to improve and automate services in the healthcare sector: namely, they can potentially improve the access to healthcare, improve doctor–patient and clinic–patient communication, or help to manage the increasing demand for health services such as via remote testing, medication adherence monitoring or teleconsultations ^[2]. Moreover, the COVID-19 global pandemic has underscored the necessity for healthcare services to be delivered in a home-based setting without compromising the safety of healthcare professionals or patients. Additionally, the pandemic has emphasized the urgency of developing new services that are both fast and tailored to the specific needs of the emerging situation. Also, the aging population and the rise of chronic illness, which cannot be treated at care facilities, demand a functional preventive care and treatment for these diseases at home. Monitoring patients at home, allowing them to follow medical prevention prescriptions, will help elderly people and people with chronic diseases stay healthy. This preventive health model will help prevent disease, detect disease early, and improve healthcare results.

Large Language Models (LLMs) have shown promising advancements in natural language processing, making them increasingly relevant for healthcare applications, particularly in the context of conversational AI. Some LLMs were already developed to create models with scientific domain-specific knowledge, such as Galactica ^[3] or BioGPT ^[4]. Nevertheless, few LLMS exist that used real-world data, such as EHR, due to the sensitivity of medical data. Two approaches developed include an accurate model, named BEHRT, used for the prediction of next/future diseases ^[5], and the GatorTron model that allows the de-identification of clinical notes ^[6]. Despite these promising applications of LLMs in healthcare, they appear ill-prepared to be deployed directly as patient tools, particularly as chatbots. This is due to the essential need for ensuring patient safety and accuracy beyond merely generating eloquent text outputs and human-like interactivity. LLMs may not be suitable in this context due to risks relating to the associations and biases of its training data. Therefore, domain-specific chatbots are more appropriate in such cases, offering tailored and focused interactions with users ^[7].

2. Chatbots in the Healthcare Industry

The idea of developing machines that can communicate with people started several decades ago. However, over time, the development of more powerful computational tools and the refinement of Natural Language Processing (NLP) techniques allowed these machines, including the chatbots, to become increasingly sophisticated in their ability to communicate effectively with users ^[8]. This happened in several fields, and healthcare was not an exception.

The first known chatbot used in healthcare was ELIZA ^[9], which was programmed in 1966 and functioned as a therapist by responding to user input with follow-up questions. ELIZA utilized the Pattern Matching and Replacement methodology to create the illusion of understanding. Another prominent chatbot, Parry ^[10], was created in 1972 to simulate a paranoid schizophrenic patient for the purpose of training psychiatrists. At a conference that same year, a demonstration featured a conversation between ELIZA and Parry ^[10]. Although ELIZA had limited knowledge and communication skills, advances in AI techniques have led to improved performance in more recent chatbots.

Chatbots have already found numerous applications in the healthcare industry. Chatbots that provide medical information, such as Healthily ^[11] and Ada Health ^[12], are some of the most widely used. Additionally, chatbots like Woebot ^[13] have been developed specifically to assist with mental health concerns and have gained significant popularity. These chatbots can offer cognitive behavioral therapy for conditions such as depression, post-traumatic stress disorder, and anxiety as well as help patients with autism improve their social skills. Some chatbots have also been designed to automate administrative tasks within healthcare systems, such as scheduling medical appointments. These chatbots can be integrated into the medical system and accessed through a user’s device to find a suitable physician and appointment slot. Other chatbots are designed to collect patient data by asking simple questions about personal information and symptoms, which can then be used for admission, symptom tracking, doctor–patient communication, and medical record keeping. Chatbots can even assist patients with requesting prescription refills, especially for chronic diseases that do not require medical intervention.

Table 1 presents a summary of the chatbots selected.

Table 1. Selected chatbots summary.

Chatbot Purpose	Chatbot Name
Diagnosis	Babylon ^[14]
	Ada Health ^[15]
	Buoy Health ^[16]
	Your.md ^[17]
	Gyant ^[18]
	Symptomate ^[19]
	MedWhat ^[20]
Well-Being	Iona Mind ^[21]
	Florence Nightingale Chatbot ^[22]
	Izzy ^[23]
	SafeDrugBot ^[24]
Mental Health Chatbots	Elomia ^[25]
Mental Health Chatbots	Youper ^[26]
Cancer-Specific Chatbots	One Remission Chatbot ^[14]
Cancer-Specific Chatbots	CancerChatbot ^[27]
Dementia-Specific Chatbots	Care ^[28]
Dementia-Specific Chatbots	AlzBot ^[29]
COVID-19-Specific Chatbots	COVID-19 Leave Chat Bot ^[30]
	MyGov Corona Helpdesk chatbot ^[31]
	Cosibot ^[30]
	ANA ^[32]
	COVIBOT ^[33]
	ChatBot-19 Risk Assessment Chatbot ^[30]

2.1. Chatbots for Diagnosis

Babylon, Ada Health, Buoy Health and Your.md are all chatbots that enable users to input their symptoms and receive helpful diagnoses ^[34]. Babylon functions by cross-referencing the symptoms provided by the user with a database of multiple diseases. This enables the user to obtain a diagnosis and make informed decisions based on the information provided. Additionally, it allows a direct interaction with a real doctor, which is not typically found in chatbots. In 2017, the United Kingdom’s National Health Service (NHS) started a trial of using this chatbot, and today, it provided over 700,000 online consultations with doctors ^[14]. Ada Health enables users to input their symptoms and then search for corresponding diseases, displaying the most probable matches ^[15]. Buoy Health is very similar to the previous chatbots. It analyzes various diseases and medical causes to propose a treatment ^[16]. Your.md is also focused on making diagnosis. In 2020, Aleksandar Ćirković ^[17] compared these four chatbots and observed that Ada Health and Your.MD showed the best results. Nevertheless, the treatment recommendations given by the chatbots for the same starting symptoms varied significantly, with Ada generally recommending emergency care for almost every diagnosis. Moreover, Ada Health asks redundant questions at the end of the consultation, such as asking about a symptom that was already mentioned earlier. Also, the additional information it provides to the users in the form of a visual description of how many patients with similar symptoms have been diagnosed with the suggested diagnosis may not be entirely accurate. Your.MD provided the most valid recommendations in the study but also raised some concerns on the dry eyes diagnosis, unnecessarily transitioning from self-treatment to emergency care.

Gyant ^[18] is another chatbot that aims to diagnosis and treat non-urgent medical conditions by asking the user questions about their symptoms and overall well-being. This information is also used to make recommendations to the healthcare professional, who will write the medical prescription ^[35]. An important feature of the Gyant chatbot is its human-like behavior due to its sense of humor, use of emojis and memes. Nevertheless, when trying to access Gyant, it is no longer available. Symptomate chatbot ^[19] allows its users to report their symptoms and to answer some questions about their health status to receive a list of the most likely diagnosis. The accuracy of the Symptomate chatbot was around 66%, which is a low result compared to, for example, ADA Health, which scored 80% ^[36]. MedWhat ^[20] is another chatbot that is useful to facilitate the medical diagnosis both for patients and doctors. It excels in speed, ease and transparency.

Well-Being Chatbots

Iona Mind ^[21] is an AI application which acts as a guide for “fitness of the mind”. It is a self-help application that promotes personalized plans and goals related to mental well-being, like managing stress and anxiety, to help users achieve their mental wellness goals. Despite not being designed to treat psychiatric disorders, 86% of Iona Mind users report feeling better after their first session. The Florence Nightingale Chatbot ^[22] aims to help English speakers manage their health and well-being, reminding them to take their medicines and helping them find health specialists. Florence also successfully tracks patients’ health to help them reach their goals. For example, it can track user’s body weight, mood or period.

Izzy ^[23] is designed to help menstruating people monitor their period through Facebook Messenger. Izzy can assist the user with information such as when to take birth control pills, when their fertility period is, and when their next period is due. Furthermore, Izzy can remember and identify the fertile periods and inform about sexual problems ^[37]. SafeDrugBot ^[24] provides information about the safety of a drug to breastfeeding women. It is useful for medical professionals who seek information about a specific medication and want to know if it is recommended for breastfeeding women ^[38].

2.2. Mental Health Chatbots

Elomia ^[25] is a chatbot focused on treating mental health diseases. It acts like a therapist, allowing the provision of primary psychotherapy care. Using AI, it provides a bunch of functionalities capable of aiding its user, such as calming exercises, sleeping exercises, exercises to fight/reduce anxiety, breathing exercises and also exercises to improve low self-esteem. Elomia fulfilled its goals insofar as it reduced depressive tendencies up to 28% and anxiety tendencies by up to 31%. Youper ^[26] is other AI-powered chatbot that aims to help combat mental health disorders, specifically anxiety and depression. It utilizes every interaction with the user to encourage positive emotions. If the chatbot detects that the user is in a state of anxiety or depression, it suggests exercises to alleviate the symptoms. Within two weeks of using the app, users reported a moderate reduction in symptoms of anxiety and depression. The app was able to maintain the decrease in anxiety symptoms over the four-week period with an effect size of 60% ^[25]. The decrease in depression symptoms was also sustained over the four-week period, with an effect size of 42%, although there was a slight increase in depression symptoms between weeks 2 and 4. Moreover, users of Youper reported high success in managing their negative emotions through the app’s conversational AI feature ^[26].

2.3. Cancer-Specific Chatbots

In the context of the fight against cancer, the One Remission Chatbot ^[14] was developed with the goal of providing individuals affected by cancer with the information they need. In English, the chatbot offers a comprehensive list of post-cancer practices, diets, and daily exercises to aid users in managing their health and reduce their dependence on doctors. For example, it allows users to search for cancer-related risks and benefits of specific foods, and it offers the option to consult a real oncologist at all times. One Remission also serves as a mental and physical health assistant, enabling patients to share thoughts and obtain accurate explanations to their questions, either verbally or through text messages, whether they need advice on diets, exercise, or sleep.

CancerChatbot ^[27] is another cancer-specific chatbot. It collects user data to provide relevant information and can list all the symptoms of any type of cancer upon request with the option of explaining them further if needed. The chatbot also offers information on cancer stages and treatments.

2.4. Dementia-Specific Chatbots

Care ^[28] is a chatbot that aims to enhance brain health and reduce the risk of dementia by promoting regular cognitive engagement and memory recall. It achieves this by offering a series of questionnaires on various topics, which are primarily focused on biographical information. The chatbot is available on Telegram, and some questions are designed with multiple-choice options to accommodate elderly individuals with limited technical skills. Users can select an answer from the provided choices, which will be saved as their response to the corresponding question. AlzBot ^[29] is a chatbot specifically designed to assist individuals with Alzheimer’s disease, which is a form of dementia. In addition to providing conversation, it also displays reminders provided by the caregiver.

2.5. COVID-19-Specific Chatbots

Most of the chatbots related with COVID-19 aim to accurately answer COVID-19 questions, preventing the spread of misinformation among the public. Some examples are Cosibot, COVID-19 Leave Chat Bot ^[30], and the MyGov Corona Helpdesk chatbot ^[31]. Cosibot has a Portuguese version, titled Helena, which is currently disabled. COVID-19 Leave Chat Bot only provides information about the Oregon state in the United States of America. The MyGov Corona Helpdesk chatbot also assists citizens in finding nearby COVID vaccine centers.

ANA ^[32] is a chatbot assistant designed for Brazilian Portuguese speakers that provides education and information on COVID-19 while also monitoring suspected cases based on patients’ symptoms and comorbidities. However, this monitoring service is currently only available for the cities of Belo Horizonte and Divinópolis in Brazil. COVIBOT ^[33] is another chatbot that helps users seek appropriate medical attention based on their symptoms and provides information on nearby hospitals. Lastly, ChatBot-19 Risk Assessment Chatbot ^[30] calculates the user’s risk of contracting COVID-19 based on information such as their symptoms, age, and gender. It is worth noting that other commonly used chatbots for diagnosis, while not specifically designed for COVID-19, are still capable of diagnosing the disease based on symptoms.

References

Maroengsit, W.; Piyakulpinyo, T.; Phonyiam, K.; Pongnumkul, S.; Chaovalit, P.; Theeramunkong, T. A Survey on Evaluation Methods for Chatbots. In Proceedings of the 2019 7th International Conference on Information and Education Technology (ICIET 2019), Aizu, Japan, 29–31 March 2019.
Nadarzynski, T.; Miles, O.; Cowie, A.; Ridge, D. Acceptability of artificial intelligence (AI)-led chatbot services in healthcare: A mixed-methods study. Digit. Health 2019, 5, 2055207619871808.
Taylor, R.; Kardas, M.; Cucurull, G.; Scialom, T.; Hartshorn, A.S.; Saravia, E.; Poulton, A.; Kerkez, V.; Stojnic, R. Galactica: A Large Language Model for Science. arXiv 2022, arXiv:2211.09085.
Luo, R.; Sun, L.; Xia, Y.; Qin, T.; Zhang, S.; Poon, H.; Liu, T. BioGPT: Generative pre-trained transformer for biomedical text generation and mining. Brief. Bioinform. 2022, 23, bbac409.
Li, Y.; Rao, S.; Solares, J.; Hassaine, A.; Ramakrishnan, R.; Canoy, D.; Zhu, Y.; Rahimi, K.; Salimi-Khorshidi, G. BEHRT: Transformer for Electronic Health Records. Sci. Rep. 2020, 10, 7155.
Yang, X.; Chen, A.; PourNejatian, N.; Shin, H.C.; Smith, K.E.; Parisien, C.; Compas, C.; Martin, C.; Flores, M.G.; Zhang, Y.; et al. GatorTron: A Large Clinical Language Model to Unlock Patient Information from Unstructured Electronic Health Records. arXiv 2022, arXiv:2203.03540.
Au Yeung, J.; Kraljevic, Z.; Luintel, A.; Balston, A.; Idowu, E.; Dobson, R.J.; Teo, J.T. AI chatbots not yet ready for clinical use. Front. Digit. Health 2023, 5.
Caldarini, G.; Jaf, S.; McGarry, K. A Literature Survey of Recent Advances in Chatbots. Information 2022, 13, 41.
Sharma, V.; Goyal, M.; Malik, D. An intelligent behaviour shown by chatbot system. Int. J. New Technol. Res. 2017, 3, 263312.
Zemčík, T. A Brief History of Chatbots. In DEStech Transactions on Computer Science and Engineering; DEStech Publishing Inc.: Lancaster, PA, USA, 2019.
Healthily Chatbot. Available online: https://www.livehealthily.com/app (accessed on 20 March 2023).
Ada Health Chatbot. Available online: https://ada.com/pt/ (accessed on 20 March 2023).
Woebot Chatbot. Available online: https://woebothealth.com/ (accessed on 20 March 2023).
Ayanouz, S.; Abdelhakim, B.A.; Benhmed, M. A smart chatbot architecture based NLP and machine learning for health care assistance. In Proceedings of the the 3rd International Conference on Networking, Information Systems & Security (NISS2020), Marrakech, Morocco, 31 March 2020–2 April 2020; pp. 1–6.
Jungmann, S.M.; Klan, T.; Kuhn, S.; Jungmann, F. Accuracy of a Chatbot (ADA) in the diagnosis of mental disorders: Comparative case study with lay and expert users. JMIR Form. Res. 2019, 3, e13863.
Liu, J. Artificial Intelligence and Data Analytics Applications in Healthcare General Review and Case Studies. In Proceedings of the CAIH2020: Proceedings of the 2020 Conference on Artificial Intelligence and Healthcare, New York, NY, USA, 23–25 October 2020; pp. 49–53.
Ćirković, A. Evaluation of Four Artificial Intelligence–Assisted Self-Diagnosis Apps on Three Diagnoses: Two-Year Follow-Up Study. J. Med. Internet Res. 2020, 22, e18097.
Dechert, M. Implementation and Evaluation of a Chatbot to Crowdsource Geotagged Images to Detect Mosquito Breeding Sites. Master’s Thesis, University of Bremen, Bremen, Germany, 2019.
Symptomate Chatbot. Available online: https://symptomate.com/about/ (accessed on 20 March 2023).
Shaikh, A.; More, D.; Puttoo, R.; Shrivastav, S.; Shinde, S. A survey paper on chatbots. J. Eng. Technol. (IRJET) 2019, 6.
Wintermeyer, S.A. Positive Psychology Apps: A Systematic Review of the Quality and Characteristics of a Selection of Current Free-of-Charge Positive Psychological Apps Aiming to Enhance Resilience Available in the Google Play Store. Master’s Thesis, University of Twente, Enschede, The Netherlands, 2021.
Chitranshi, S.; Daml, P. Intelligent Interaction via Chatbots Transcending the Current Constructs. PalArch’s J. Archaeol. Egypt/Egyptol. 2020, 17, 4568–4576.
Piumali, K.; Hettige, B. A Review on Existing Health Care Monitoring Chatbots. 2021. Available online: https://www.researchgate.net/profile/Hansika-Piumali/publication/357929804_A_Review_on_Existing_Health_Care_Monitoring_Chatbots/links/61e7ed149a753545e2e0c281/A-Review-on-Existing-Health-Care-Monitoring-Chatbots.pdf (accessed on 20 May 2023).
Poola, I. Making artificial intelligence (AI) and disrupted business intelligence (BI) truly conversational with humanity touch, automated descriptions and talking bots. J. Adv. Res. Ideas Innov. Technol. 2017, 3, 573–577.
Romanovskyi, O.; Pidbutska, N.; Knysh, A. Elomia Chatbot: The Effectiveness of Artificial Intelligence in the Fight for Mental Health. In Proceedings of the COLINS, Samorin, Slovakia, 28 August 2021; pp. 1215–1224.
Mehta, A.; Niles, A.N.; Vargas, J.H.; Marafon, T.; Couto, D.D.; Gross, J.J. Acceptability and effectiveness of artificial intelligence therapy for anxiety and depression (Youper): Longitudinal observational study. J. Med. Internet Res. 2021, 23, e26771.
Belfin, R.; Shobana, A.; Manilal, M.; Mathew, A.A.; Babu, B. A graph based chatbot for cancer patients. In Proceedings of the 2019 ICACCS, Coimbatore, India, 15–16 March 2019; IEEE: New York, NY, USA, 2019; pp. 717–721.
Müller, C.; Paluch, R.; Hasanat, A. Care: A chatbot for dementia care. In Proceedings of the Mensch und Computer 2022-Workshopband, Darmstadt, Germany, 4–7 September 2022.
Le Xin, T.; Arshad, A.; bin Abdul Salam, Z.A. AlzBot-Mobile App Chatbot for Alzheimer’s Patient to be Active with Their Minds. In Proceedings of the 14th International Conference on the Developments on eSystems Engineering (DeSE2021), Sharjah, United Arab Emirates, 7–10 December; IEEE: New York, NY, USA, 2021; pp. 124–129.
Carvalho, R.M.S. Chatbot de Informação Sobre o COVID-19. Ph.D. Thesis, Universidade de Coimbra, Coimbra, Portugal, 2021.
Padmavathi, S. Artificial Intelligence in Health Care. Ann. SBV 2022, 10, 23.
Marcolino, T.C.F.M.S.; Prates, I.R.R.; Adriana, L.B.R.Z.R.; Meira, P.W., Jr.; Ribeiro, A.L. ANA: A Brazilian Chatbot Assistant about COVID-19. Available online: https://www.researchgate.net/profile/Zilma-Reis/publication/342869406_Ana_a_brazilian_chatbot_assistant_about_covid_19/links/5f0e561b92851c1eff11b3d5/Ana-a-brazilian-chatbot-assistant-about-covid-19.pdf (accessed on 1 January 2023).
Khan, N.A.; Albatein, J. COVIBOT-An intelligent WhatsApp based advising bot for Covid-19. In Proceedings of the 2021 ICCIKE, Dubai, United Arab Emirates, 17–18 March 2021; IEEE: New York, NY, USA, 2021; pp. 418–422.
Wang, W.; Siau, K. Trust in Health Chatbots. 2018. Available online: https://www.researchgate.net/profile/Keng-Siau-2/publication/333296446_Trust_in_Health_Chatbots/links/5ce5dd35299bf14d95b1d15b/Trust-in-Health-Chatbots.pdf (accessed on 12 January 2023).
Makasi, T.; Nili, A.; Desouza, K.; Tate, M. Chatbot-mediated public service delivery: A public service value-based framework. First Monday 2020, 25.
Swick, R.K. The Accuracy of Artificial Intelligence (AI) Chatbots in Telemedicine. J. South Carol. Acad. Sci. 2021, 19, 17.
Carvalho, L.; Albuquerque, M.; Nogueira, R.; Lochter, J. Dra. Lara: Assistente Virtual de Apoio e Acompanhamento ao Pré-Natal. 2019. Available online: https://www.researchgate.net/profile/Cristiana-De-Muylder/publication/338665851_SOCIAL_CRM_E_FINTECH_ANALISE_DE_PUBLICACOES_INTERNACIONAIS_2008-2018/links/5fca526da6fdcc697bdc0711/SOCIAL-CRM-E-FINTECH-ANALISE-DE-PUBLICACOES-INTERNACIONAIS-2008-2018.pdf#page=276 (accessed on 15 January 2023).
Bates, M. Health care chatbots are here to help. IEEE Pulse 2019, 10, 12–14.

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.

Upload a video for this entry

Information

Subjects: Computer Science, Artificial Intelligence

Contributors MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :

Eva Maia

Pedro Vieira

Isabel Praça

View Times: 213

Update Date: 24 Oct 2023

Table of Contents

Video Upload Options

Confirm