Academy of Diagnostics & Laboratory Medicine - Scientific Short
Scientific Shorts are brought to you by the
Artificial intelligence (AI) conversational agents, also known as chatbots or intelligent virtual assistants, are computer programs designed to simulate human conversation and use natural language processing to interact with users through messaging applications, websites, or mobile apps (1). In recent years, AI chatbots are finding their way into an expanding range of healthcare domains, offering diverse functions and applications. The COVID-19 pandemic has accelerated the adoption and deployment of chatbots, which were designed to disseminate health information and knowledge about symptoms, precautionary measures, and medical treatment (2). In addition, AI chatbots, such as Babylon Health and Ada Health, have been used to monitor patients with specific conditions, provide patients with information and personalized risk assessment, support healthcare delivery, and facilitate workflow automation. Chatbots have been found to offer potential utility in a variety of clinical areas, such as diabetes, obesity, depression, anxiety, substance use disorders, smoking cessation, and cancer.
In the realm of Laboratory Medicine, chatbots hold great potential for improving the automation of responses to routine inquiries concerning laboratory tests, as well as to facilitate the interpretation of laboratory results. Many clinicians are perplexed by the complexity of laboratory tests and struggle to find detailed testing instructions. They sometimes interpret test results without a comprehensive understanding of the instrument platform, principles, and clinical indication of various laboratory tests. Such analytical information of laboratory tests may not be easily accessible or understandable by non-laboratory personnel. However, clinicians often lack the time to thoroughly search the literature or request a laboratory consultation. As a result, clinicians need an accurate, understandable, rapid, and interactive approach to obtain information about laboratory tests and interpret laboratory results. Chatbots that could provide “curbside consultation”, as described by Lee et al. (3), have the potential to support clinicians who require immediate answers. Furthermore, many patients use online resources to research their laboratory test results and understand their medical conditions. Consequently, they are likely to seek chatbots’ assistance in interpreting laboratory data or understanding how to utilize clinical laboratory services.
The use of chatbots in the field of Laboratory Medicine is still in its infancy and fraught with risks that must be considered (4). There are several reasons for this limited use, which include a lack of domain knowledge training and expert annotation. The training data should encompass specialized medical language and terminology, which, without a deep understanding of medical practice, can be easily mistranslated or misinterpreted. Furthermore, there is a lack of robust evaluation and validation of chatbots in laboratory medicine, making it difficult to determine their accuracy and reliability. Concerns have been raised that chatbots may generate responses that appear convincing but contain inaccuracies and fabricated information, often referred to as “hallucinations” (5). Moreover, large language models (LLM), such as the Chat Generative Pre-trained Transformer (ChatGPT), are not able to reliably interpret laboratory results in clinical context, as these often require specialized knowledge to understand (5). Therefore, it is imperative for clinicians and laboratorians to understand the benefits and limitations of chatbots in order to use them effectively in their clinical practice.
AI chatbots, such as ChatGPT, are rapidly advancing AI applications at a remarkable pace, with exciting potential to enhance communication among patients, clinicians, and laboratorians. However, the current chatbots do not display adequate accuracy for clinical deployment and cannot reliably interpret complex laboratory data. Tackling AI hallucinations perhaps is the most critical challenge when using an end-to-end solution such as ChatGPT. The evidence accumulated so far underscores the necessity for extensive training of chatbots on rigorously validated medical knowledge and clinical resources, as well as comprehensive evaluation against standard clinical practice. Before endorsing their clinical use, AI chatbots must demonstrate not only enhanced accuracy but also further development to consistently provide verifiable rationale behind their response to clinical questions.
Scientific Shorts are brought to you by the
Fellows of the Academy use the designation of FADLM. This designation is equivalent to FACB and FAACC, the previous designations used by fellows of the National Academy of Clinical Biochemistry and AACC Academy. Those groups were rebranded as Academy of Diagnostics & Laboratory Medicine in 2023.
