“Hi Caroline, this is Jeff! You reported feeling heavily symptomatic yesterday. I wanted to see if you are doing better. On a scale from 1-10, how would you rate your pain today?”
No, that isn’t a doctor speaking. That’s an artificial intelligence-driven (AI) chatbot, which is providing patients new approaches to receiving the right care at their point of need. Chatbots take on a variety of forms. Their tasks vary and may include automating manual functions for both health care providers (HCPs) and patients, providing 24/7 care, offering emotional and physical wellbeing support, increasing treatment adherence, and more.
For HCPs and patients, this is a win-win. Patient-driven outcomes are the goal, and AI chatbots can vastly improve HCP functions and patient engagement, and ultimately, financial and healthcare results.
Patient Engagement Issues: Medical NONADHERENCE
Low patient engagement can cause medication nonadherence, which poses significant risks:
“Medication nonadherence for patients with chronic diseases is extremely common, affecting as many as 40% to 50% of patients who are prescribed medications for management of chronic conditions such as diabetes or hypertension. This nonadherence to prescribed treatment is thought to cause at least 100,000 preventable deaths and $100 billion in preventable medical costs per year.”
- Fred Kleinsinger, MD.¹
With Google and WebMD at arms-length away, it is crucial for treatment adherence that the healthcare industry continues advancing with chatbots, which may serve both HCPs and patients as convenient information highways and communication vehicles.
AI Chatbots: Explained
An AI chatbot does not replace an HCP. Instead, it provides additional support for both the HCP and the patient for efficiency and areas that lack thereof.
These chatbots simulate human conversation through voice commands, text chats, or both. However, unlike a human, they are available at all hours of the day to any geographic location. Also, bots are more financially sound – they cost virtually nothing to speak with and reduce labor costs for 24/7 workers.
Some chatbots are preloaded with frequently asked questions (FAQs) and answers and are limited to their programmed responses. Others are machine learning (ML) powered, which use internal neural networks to self-teach new replies, and to improve inquiry response time and accuracy.²
Chatbots can be as generic as an FAQ, and as intelligent as Amazon Alexa. Some are even on their way to inhabiting physical forms.
AI Chatbot Benefits for Patients and HCPS
manual processes to
improve HCP care.
They automate manual processes to improve HCP care.
For patients with chronic conditions, quick response times from HCPs can be critical. For this reason, chatbots help improve physician efficiency and prevent burnout.
Much like Alexa, HCPs can speak to their devices to:
- Record medical notes,
- Schedule appointments,
- Order prescriptions to specific pharmacies,
- Generate reports,
- Access information via the EHR more efficiently.
One such app is SafedrugBot, a messaging service that offers HCPs information for breastfeeding patients via the Telegram app. SafedrugBot cuts HCP studying time, providing them accurate, updated information on the medications compatible with the pregnant patient’s state. The bot also suggests suitable alternatives to certain drugs, saving up their time spent collating that information.
HCP task-automating chatbots like SafedrugBot are highly beneficial as they reserve an HCPs’ energy for higher-level tasks such as building client relationships.
They help patients identify health red flags.
Searching cold symptoms on WebMD can send the average, level-headed medical advice seeker spiraling into diagnostic paranoia. However, new-wave diagnostic chatbots provide a tailor-made alternative. These bots inquire about the user’s condition and compare it to databases of medical information and similar patient experiences to find a more personalized – and hopefully, less mortifying – response.
Ada is an app that puts the doctor in your palm. This personal health companion uses AI to determine your symptoms by asking relevant questions such as, “where does it hurt?” Then, Ada suggests possible diagnoses based on others’ responses.
The AI feature self-improves as it learns more about the user. Treatment recommendations range from at-home treatment to urging professional help. For professional help scenarios, Ada generates a report that patients can give to their HCPs to streamline procedures. Once the user deletes the app, the app eliminates the patient’s information for confidentiality.
They keep patients on-track with treatment and medical plans.
For patients, chatbot assistants can help:
- Schedule follow-ups,
- Provide daily reminders to take medicine,
- Ease symptom reporting,
- Increase payment-processing speeds,
- Answer health-related questions.
LifeLink powers chatbot technology that eases the patient experience before, at, and after the point of care. During clinical visits, their bots provide users with accurate wait times, updates on care procedures, test preparation, assistance locating doctors and making appointments, and even send loved ones live patient updates. Post-treatment, the bot informs patients about lab tests, provides specific electronic medical record (EMR) data, and referral info for treatment plans and easy appointment tracking. LifeLink even offers interactive demos online so people can experience what their chatbots can offer.
Such apps increase patient engagement by actively involving them and their loved ones in their treatment. In doing so, the patient’s uncertainty regarding their treatment diminishes, which can reduce frustration and increase overall satisfaction.
They offer a bias-free space for patients to disclose personal information.
AI chatbots, such as therapy bots, are offering an alternative to users to chat about their health.
While these bots do not replace face-to-face empathy, they can:
- Answer questions 24/7 tirelessly in a soothing, cheery tone without getting impatient,
- Provide patients with a sense of anonymity, allowing them to open about their health,
- Speak subjectively and consistently to all the patients without biases or personal issues affecting their performance.
One such chatbot, Woebot, tries to learn about the patient’s mental health and energy levels by asking questions. From there, it checks in daily with the patients and makes suggestions if any changes spark concern. In a Stanford University study, students aged 18-28 who spoke with Woebot reported significant reductions in depression and anxiety symptoms after two weeks of usage.³
Other responsive therapy bots are physical. The PARO Therapeutic Robot is a robot baby harbor seal designed to reduce human stress. This adorable, fuzzy robot animal therapy alternative is proven to elicit similar responses to live animal therapy, such as responding to stimuli (e.g., enjoying being pet and stroked). They even develop personalities that their users enjoy by learning their responses over time. For patients with dementia who may become erratic, PARO is a healthy alternative to sedation or even physical restraint.
There are many concerns that people will substitute therapy for these bots. However, such bots offer a good stand-in for those needing assistance in a pinch.
The Future of AI Chatbots in Healthcare
Today, chatbots are at the forefront of business-to-consumer communication. In 2019, the chatbot industry reached a market value of $17.17 billion and is projected to reach $102.29 billion by 2025, and the healthcare industry will certainly take heed through advancements.⁴
AI chatbots may very well inhabit bodies. Moxi, a fully-functional AI-powered robot nurse, already exists in Texas and can assist nurses with time-consuming, non-patient facing roles like fetching items. Oh yeah - Moxi
Mobilization Robots: Currently, Honda’s Hoko Assist robot helps people disabled from the waist-down physically walk upright, whereas they would have to use a wheelchair otherwise. In the future, such robots that compensate for body parts may become mainstream and overtake seated assistants such as wheelchairs.
Staff-Assisting Robots: Health care center bots may gain enough empathy to be able to instruct patients through physical therapy, help them with bedside care, and other tasks where staff might be short.
AI Chatbot Advancements
Gamification: Chatbots may use health and treatment-related games, and reward patients with points for treatment adherence.
Understanding New Disease States: ML-enabled chatbots can help HCPs with understanding new treatments instantly. The computer will sift through relevant data to make suggestions. On the patient front, they will be able to identify the top HCPs for given diseases and receive vastly improved stand-in advice when help is unavailable.
Given the technology’s potential healthcare industry usage and cost-effectiveness, the healthcare industry will continue to advance and hone this new technology. Considering the humanism of existing chatbots, a vast improvement in their abilities to mimic humans in behavior and services is in the future forecast.
- Kleinsinger F., “The Unmet Challenge of Medication Nonadherence.” The Permanente Journal. Retrieved from: https://www.researchgate.net/publication/326214247_The_Unmet_Challenge_of_Medication_Nonadherence
- Frankenfield, F., “Chatbot.” Investopedia.com. Updated June 26, 2019, Retrieved from: https://www.investopedia.com/terms/c/chatbot.asp
- Fitzpatrick KK, Darcy A, Vierhile M., “Delivering Cognitive Behavior Therapy to Young Adults With Symptoms of Depression and Anxiety Using a Fully Automated Conversational Agent (Woebot): A Randomized Controlled Trial.” JMIR Ment Health, 2017. Retrieved from: https://mental.jmir.org/2017/2/e19/
- “Chatbot Market – Growth, Trends, and Forecast (2020 – 2025),” Mordor Intelligence. 2019. Retrieved from: https://www.mordorintelligence.com/industry-reports/chatbot-market
- Phillips, C., “Natural Language Processing (NLP) & Why Chatbots Need it,” Chatbots Magazine by Octane AI, May 24, 2018. Retrieved from: https://chatbotsmagazine.com/natural-language-processing-nlp-why-chatbots-need-it-a9d98f30ab13