Health-focused conversational agents in person-centered care: a review of apps npj Digital Medicine

Patients can quickly assess symptoms and determine their severity through healthcare chatbots that are trained to analyze them against specific parameters. However, healthcare providers may not always be available to attend to every need around the clock. This is where chatbots come into play, as they can be accessed by anyone at any time.

For each app, data on the number of downloads were abstracted for five countries with the highest numbers of downloads over the previous 30 days. Chatbot apps were downloaded globally, including in several African and Asian countries with more limited smartphone penetration. The United States had the highest number of total downloads (~1.9 million downloads, 12 apps), followed by India (~1.4 million downloads, 13 apps) and the Philippines (~1.25 million downloads, 4 apps). Details on the number of downloads and app across the 33 countries are available in Appendix 2. Only ten apps (12%) stated that they were HIPAA compliant, and three (4%) were Child Online Privacy and Protection Act (COPPA)-compliant.

It is critical to incorporate multilingual support and guarantee accessibility in order to serve a varied patient population. By taking this step, the chatbot’s reach is increased and it can effectively communicate with users who might prefer a different language or who need accessibility features. Additionally, we offer consulting services to explore how best to use AI technology in your own patient communication software applications.

Chatbots in treatment

Disruptions due to the pandemic affected people with chronic conditions who could not access routine medical services. It let to postponed elective procedures, on-site visits, and reduced rates of hospitalization during the COVID-19 emergency. Artificial intelligence–driven voice technology deployed on mobile phones and smart speakers has the potential to improve patient management and organizational workflow. Voice chatbots have been already implemented in health care–leveraging innovative telehealth solutions during the COVID-19 pandemic. They allow for automatic acute care triaging and chronic disease management, including remote monitoring, preventive care, patient intake, and referral assistance.

Personalization features were only identified in 47 apps (60%), of which all required information drawn from users’ active participation. Forty-three of these (90%) apps personalized the content, and five (10%) personalized the user interface of the app. Examples of individuated content include the healthbot asking for the user’s name and addressing them by their name; or the healthbot asking for the user’s health condition and providing information pertinent to their health status.

Chatbots were found to have improved medical service provision by reducing screening times [17] and triaging people with COVID-19 symptoms to direct them toward testing if required. These studies clearly indicate that chatbots were an effective tool for coping with the large numbers of people in the early stages of the COVID-19 pandemic. Overall, this result suggests that although chatbots can achieve useful scalability properties (handling many cases), accuracy is of active concern, and their deployment needs to be evidence-based [23].

As well, virtual nurses can send daily reminders about the medicine intake, ask patients about their overall well-being, and add new information to the patient’s card. In this way, a patient does not need to directly contact a doctor for an advice and gains more control over their treatment and well-being. And due to a fact that the bot is basically a robot, all these actions take little time and the appointment can be scheduled within minutes.

More research is needed to fully understand the effectiveness of using chatbots in public health. Concerns with the clinical, legal, and ethical aspects of the use of chatbots for health care are well founded given the speed with which they have been adopted in practice. Future research on their use should address these concerns through the development of expertise and best practices specific to public health, including a greater focus on user experience.

From the emergence of the first chatbot, ELIZA, developed by Joseph Weizenbaum (1966), chatbots have been trying to ‘mimic human behaviour in a text-based conversation’ (Shum et al. 2018, p. 10; Abd-Alrazaq et al. 2020). Thus, their key feature is language and speech recognition, that is, natural language processing (NLP), which enables them to understand, to a certain extent, the language of the user (Gentner et al. 2020, p. 2). Hesitancy from physicians and poor adoption by patients is a major barrier to overcome, which could be explained by many of the factors discussed in this section. A cross-sectional web-based survey of 100 practicing physicians gathered the perceptions of chatbots in health care [6]. Although a wide variety of beneficial aspects were reported (ie, management of health and administration), an equal number of concerns were present. Over 70% of physicians believe that chatbots cannot effectively care for all the patients’ needs, cannot display human emotion, cannot provide detailed treatment plans, and pose a risk if patients self-diagnose or do not fully comprehend their diagnosis.

Chatbots must be designed with the user in mind, providing patients a seamless and intuitive experience. Healthcare providers can overcome this challenge by working with experienced UX designers and testing chatbots with diverse patients to ensure that they meet their needs and chatbot technology in healthcare expectations. Telemedicine uses technology to provide healthcare services remotely, while chatbots are AI-powered virtual assistants that provide personalized patient support. They offer a powerful combination to improve patient outcomes and streamline healthcare delivery.

EXPERT-RECOMMENDED AI CHATBOT IDEAS

These chatbots can provide live feedback to help patients get an overview of their symptoms, become aware of their illness, triage and manage their conditions, and ultimately improve their health [19-22]. Such chatbots act as a virtual conversational agent mimicking human interactions and offering medical advice (eg, diagnostic suggestions) directly to patients in a timely and cost-effective manner. In this way, health chatbots provide a form of triage into the health care system and become the first point of contact for health.

• Two-thirds (21/32, 66%) of the chatbots in the included studies were developed on custom-developed platforms on the web [6,16,20-26], for mobile devices [21,27-36], or personal computers [37,38].
• When physicians observe a patient presenting with specific signs and symptoms, they assess the subjective probability of the diagnosis.
• Also, she says, “it is imperative that what’s available to the public is clinically and rigorously tested,” she says.
• These findings highlight the importance of providing more useful information that patients need.

They expect that algorithms can make more objective, robust and evidence-based clinical decisions (in terms of diagnosis, prognosis or treatment recommendations) compared to human healthcare providers (HCP) (Morley et al. 2019). Thus, chatbot platforms seek to automate some aspects of professional decision-making by systematising the traditional analytics of decision-making techniques (Snow 2019). In the long run, algorithmic solutions are expected to optimise the work tasks of medical doctors in terms of diagnostics and replace the routine tasks of nurses through online consultations and digital assistance. In addition, the development of algorithmic systems for health services requires a great deal of human resources, for instance, experts of data analytics whose work also needs to be publicly funded. A complete system also requires a ‘back-up system’ or practices that imply increased costs and the emergence of new problems. The crucial question that policy-makers are faced with is what kind of health services can be automated and translated into machine readable form.

Voice chatbots can potentially help patients easily communicate their health status by providing them with any disease management data. This approach allows for remote monitoring of medical patients without COVID-19 and those with COVID-19 who are mildly ill. The implementation of RPA technology integrating medical data collected through a conversational interface with the hospital database and alert-based CDSS delivers a powerful architecture that can function hand-in-hand with health care providers. Automatic clinical follow-up services provide access to up-to-date information about the individual’s health status for informed medical decision-making [57] and reduce the risk of exposure and infection during face-to-face contact.

• All authors contributed to the assessment of the apps, and to writing of the manuscript.
• Furthermore, Rasa also allows for encryption and safeguarding all data transition between its NLU engines and dialogue management engines to optimize data security.
• Algorithms are still not at a point where they can mimic the complexities of human emotion, let alone emulate empathetic care, she says.
• With the advances in artificial intelligence (AI) technology in recent years, there is an opportunity to tackle the challenges and barriers faced by patients in seeking timely health information and to reduce the burdens posed on medical professionals [16,17].
• This requires the same kind of plasticity from conversations as that between human beings.

Bella, one of the most advanced text-based chatbots on the market advertised as a coach for adults, gets stuck when responses are not prompted [51]. Given all the uncertainties, chatbots hold potential for those looking to quit smoking, as they prove to be more acceptable for users when dealing with stigmatized health issues compared with general practitioners [7]. For example, IBM’s Watson for Oncology examines data from records and medical notes to generate an evidence-based treatment plan for oncologists [34]. Studies have shown that Watson for Oncology still cannot replace experts at this moment, as quite a few cases are not consistent with experts (approximately 73% concordant) [67,68]. Nonetheless, this could be an effective decision-making tool for cancer therapy to standardize treatments.

With the rapidly increasing applications of chatbots in health care, this section will explore several areas of development and innovation in cancer care. Various examples of current chatbots provided below will illustrate their ability to tackle the triple aim of health care. The specific use case of chatbots in oncology with examples of actual products and proposed designs are outlined in Table 1.

The exponentially increasing number of patients with cancer each year may be because of a combination of carcinogens in the environment and improved quality of care. The latter aspect could explain why cancer is slowly becoming a chronic disease that is manageable over time [19]. Added life expectancy poses new challenges for both patients and the health care team. For example, many patients now require extended at-home support and monitoring, whereas health care workers deal with an increased workload. Although clinicians’ knowledge base in the use of scientific evidence to guide decision-making has expanded, there are still many other facets to the quality of care that has yet to catch up. Key areas of focus are safety, effectiveness, timeliness, efficiency, equitability, and patient-centered care [20].

Also, approximately 89% of healthcare organizations state that they experienced an average of 43 cyberattacks per year, which is almost one attack every week. We are dedicated to providing cutting-edge healthcare software solutions that improve patient outcomes and streamline healthcare processes. Evolving into versatile educational instruments, chatbots deliver accurate and relevant health information to patients. This empowerment enables individuals to make well-informed decisions about their health, contributing to a more health-conscious society. Consider KMS Healthcare as your go-to resource for the development and consulting expertise you need to explore how you can use AI to improve patient communication software applications. Gathering user feedback is essential to understand how well your chatbot is performing and whether it meets user demands.

The American Medical Association has also adopted the Augmented Intelligence in Health Care policy for the appropriate integration of AI into health care by emphasizing the design approach and enhancement of human intelligence [109]. An area of concern is that chatbots are not covered under the Health Insurance Portability and Accountability Act; therefore, users’ data may be unknowingly sold, traded, and marketed by companies [110]. On the other hand, overregulation may diminish the value of chatbots and decrease the freedom for innovators.

The Future of Chatbots in Healthcare Settings

Before designing a conversational pathway for an AI driven healthcare bot, one must first understand what makes a productive conversation. Before chatbots, we had text messages that provided a convenient interface for communicating with friends, loved ones, and business partners. In fact, the survey findings reveal that more than 82 percent of people keep their messaging notifications on. Once the fastest-growing health app in Europe, Ada Health has attracted more than 1.5 million users, who use it as a standard diagnostic tool to provide a detailed assessment of their health based on the symptoms they input. Any chatbot you develop that aims to give medical advice should deeply consider the regulations that govern it. There are things you can and cannot say, and there are regulations on how you can say things.

First, it can perform an assessment of a health problem or symptoms and, second, more general assessments of health and well-being. Third, it can perform an ‘assessment of a sickness or its risks’ and guide ‘the resident to receive treatment in services promoting health and well-being within Omaolo and in social and health services external to’ it (THL 2020, p. 14). In the aftermath of COVID-19, Omaolo was updated to include ‘Coronavirus symptoms checker’, a service that ‘gives guidance regarding exposure to and symptoms of COVID-19’ (Atique et al. 2020, p. 2464; Tiirinki et al. 2020). 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). Many experts have emphasised that chatbots are not sufficiently mature to be able to technically diagnose patient conditions or replace the judgements of health professionals.

Chatbots can extract patient information by asking simple questions such as their name, address, symptoms, current doctor, and insurance details. The chatbots then, through EDI, store this information in the medical facility database to facilitate patient admission, symptom tracking, doctor-patient communication, and medical record keeping. Conversational chatbots with different intelligence levels can understand the questions of the user and provide answers based on pre-defined labels in the training data. Finally, the issue of fairness arises with algorithm bias when data used to train and test chatbots do not accurately reflect the people they represent [101]. As the AI field lacks diversity, bias at the level of the algorithm and modeling choices may be overlooked by developers [102].

However, the field of chatbot research is in its infancy, and the evidence for the efficacy of chatbots for prevention and intervention across all domains is at present limited. Surprisingly, there is no obvious correlation between application domains, chatbot purpose, and mode of communication (see Multimedia Appendix 2 [6,8,9,16-18,20-45]). Some studies did indicate that the use of natural language was not a necessity for a positive conversational user experience, especially for symptom-checking agents that are deployed to automate form filling [8,46].

Few of the included studies discussed how they handled safeguarding issues, even if only at the design stage. This methodology is a particular concern when chatbots are used at scale or in sensitive situations such as mental health. In this respect, chatbots may be best suited as supplements to be used alongside existing medical practice rather than as replacements [21,33]. The timeline for the studies, illustrated in Figure 3, is not surprising given the huge upsurge of interest in chatbots from 2016 onward. Although health services generally have lagged behind other sectors in the uptake and use of chatbots, there has been greater interest in application domains such as mental health since 2016. This finding may reflect both the degree to which conversational technologies lend themselves to the kinds of interactive methodologies used in mental health and the necessity for greater scrutiny of the methods that are used by health practitioners in field.

To seamlessly implement chatbots in healthcare systems, a phased approach is crucial. Start by defining specific objectives for the chatbot, such as appointment scheduling or symptom checking, aligning with existing workflows. Identify the target audience and potential user scenarios to tailor the chatbot’s functionalities. Integration with electronic health record (EHR) systems streamlines access to relevant patient data, enhancing personalized assistance. Regularly update the chatbot based on user feedback and healthcare advancements to ensure continuous alignment with evolving workflows.

Woman uses AI chatbot for mental health support, says it is more convenient than visiting a therapist

The development—especially conceptual in nature—of ADM has one of its key moments in the aftermath of World War II, that is, the era of the Cold War. America and the Soviets were both keen (in their own ways) on find ways to automatise and streamline their societies (including decision-making). This was led by people from different fields of science, who were reconceptualising human reason ‘as rationality’ (p. 29), thus creating formal models of functions and processes of biological and artificial organisms, firms, organisations and even societies. In the field of medical practice, probability assessments has been a recurring theme.

An effective UI aims to bring chatbot interactions to a natural conversation as close as possible. And this involves arranging design elements in simple patterns to make navigation easy and comfortable. All these platforms, except for Slack, provide a Quick Reply as a suggested action that disappears once clicked.

Depending on their type (more on that below), chatbots can not only provide information but automate certain tasks, like review of insurance claims, evaluation of test results, or appointments scheduling and notifications. You can foun additiona information about ai customer service and artificial intelligence and NLP. By having a smart bot perform these tedious tasks, medical professionals have more time to focus on more critical issues, which ultimately results in better patient care. While a chatbot in healthcare can not be considered a 100% trusted and reliable medical consultant, it can at least help patients recognize their symptoms and the urgency of their condition or answer their questions.

Conversely, automation errors have a negative effect on trust—‘more so than do similar errors from human experts’ (p. 25). However, the details of experiencing chatbots and their expertise as trustworthy are a complex matter. As Nordheim et al. have pointed out, ‘the answers not only have to be correct, but they also need to adequately fulfil the users’ needs and expectations for a good answer’ (p. 25). Importantly, in addition to human-like answers, the perceived human-likeness of chatbots in general can be considered ‘as a likely predictor of users’ trust in chatbots’ (p. 25).

It is important to consider continuous learning and development when developing healthcare chatbots. The health bot uses machine learning algorithms to adapt to new data, expanding medical knowledge, and changing user needs. In the first stage, a comprehensive needs analysis is conducted to pinpoint particular healthcare domains that stand to gain from a conversational AI solution.

Healthcare chatbots offer a convenient and accessible way for patients to access healthcare information, receive support, and manage their health remotely. Chatbots typically ask consecutive questions about concomitant symptoms so that they can generate a more accurate diagnosis. However, these questions are usually hard to answer and can easily overwhelm the user.

Service-provided classification is dependent on sentimental proximity to the user and the amount of intimate interaction dependent on the task performed. This can be further divided into interpersonal for providing services to transmit information, intrapersonal for companionship or personal support to humans, and interagent to communicate with other chatbots [14]. The next classification is based on goals with the aim of achievement, subdivided into informative, conversational, and task based. Response generation chatbots, further classified as rule based, retrieval based, and generative, account for the process of analyzing inputs and generating responses [16].