Marketers from all industries have always relied on outbound campaigns to reach potential customers. Before the time of the Internet, letters and phone calls were the main means to reach customers. 

Then email blast campaigns became popular, as they offer a cheaper and more efficient way to reach out to customers. Email campaigns evolved into email drip campaigns, where a marketer can design a sequence of emails that will be sent over time to leads. Conditional drip campaigns are also possible, where the next email depends on whether the lead opened a previous email.

Email campaigns suffer from low open rates. Further, they are unable to effectively complete a conversion. For example, the email may encourage a patient to call the doctor to schedule an appointment or send a customer to a website to shop for concert tickets.  

After email campaigns, came texting campaigns. The key promise compared to email campaigns is higher open rates. A marketer needs to have a list of phone numbers instead of a list of emails. The message of course has to be shorter, typically an image  plus a small paragraph of text, possibly with emoticons, and a hyperlink, to be sent as one or more MMS messages. 

Texting campaigns indeed increase the open rate, however they still suffer from the second drawback of email campaigns: they cannot complete a conversion. Instead, they typically send the lead to a website to complete the transaction. This incurs unnecessary friction and lack of instant gratification. For example, the text message may encourage a user to make an appointment, but what if the user has a question on insurance coverage or cost?

Chatbot campaigns are the evolution of texting campaigns, where a single message is released by a chatbot. That is, instead of blast sending a message to thousands of leads, we blast trigger a chatbot to the leads. For example, a texting campaign could send out message “It is time for your dental cleaning. To schedule your cleaning, go to https://mydentist.com/cleaning”. Instead, a chatbot campaign could trigger a chatbot that initially sends out message “Hi Bob, it’s been a while since your last cleaning. Is now a good time to chat about it? Yes/No.” If the user replies “No” the chatbot will try again the next day, else the chatbot will continue with message “Great, let’s get you scheduled. Has your insurance coverage changed since your last visit? Yes/No”, and the chatbot can then continue to collect all needed information and find a suitable time slot. This increases the conversion rates, improves user experience and decreases customer service cost (saves calling center time). 

How about HIPAA?

I cannot provide legal advice, but my understanding is that protected health information (PHI) can be shared over an unencrypted medium like SMS when the patient consents to it explicitly or implicitly (by initiating communication via SMS).

A second approach is to initiate communication via SMS and send a hyperlink that seamlessly takes the patient to a secure HIPAA-compliant Web-based chat. For example, the SMS message may say something like “Good morning from Dr. Smith’s office. To securely collect your information please follow this link. https://drsmith.com/chat” The user clicks on the link and a full-screen web chat starts. All interaction is associated with the phone number of the user (patient).

SmartBot360 is a leader in chatbot campaigns. It allows users to easily visually create chatbot flows and then create conditional drip campaigns around these chatbots. To read more details about how SmartBot360 supports texting campaigns, read the relevant documentation page.

There has been great publicity recently on the capabilities of ChatGPT. It is considered the state-of-the-art in conversational AI, given its large training knowledge and its strong natural language understanding. One may ask ChatGPT encyclopedic questions such as “What is the population of the USA?” or domain-specific questions such as “What is the cause of diabetes?” An interesting feature is that it can change the language style of its response, for example by asking “Explain the causes of diabetes to a small child.”

Given the powerful capabilities of ChatGPT, people have been trying to find its potential killer applications in all industries. For example, many people view it as a competitor to Google or to Alexa or Siri. 

In this article, we will discuss how ChatGPT can be used in healthcare applications. Let’s first start with a review of healthcare applications for chatbots in general, which we discussed in a previous article. For each application, we will discuss if and how ChatGPT can be leveraged.

1. Self-diagnosis

This is an area that was very hot a few years ago, and has since cooled down a little bit. There are several platforms that allow patients to enter their symptoms, such as Babylon Health,  Healthily or WebMD. These platforms use a combination of machine learning and manual expert input.  They are programmed to ask the right next question given the patient input so far. This is something that ChatGPT is not good at, that is, ChatGPT is not good at asking clarification questions to solicit the right additional information from a user. Further, it is hard to compete with algorithms trained with extensive human input and calibration in a specialized domain. 

To avoid possible liability or confusion, ChatGPT is explicit on not being able to do symptom checking, responding “I am an AI language model and do not have the ability to diagnose medical conditions. It is recommended that you consult a doctor or a medical professional for proper diagnosis and treatment.” 

Verdict: Not applicable

2. Healthcare navigation

The goal here is to help patients navigate the complex healthcare landscape; for example, find a doctor in their network, compare prices, or learn about their benefits. These chatbots are typically available only to members of a healthcare system, so they are not freely accessible on the web. The technology of these bots can range from simple multiple-choice diagrams to deep learning models that are trained on existing customer service chats.

ChatGPT is not trained to provide local information, such as look for doctors in an area. Instead, it forwards the patient to other resources: “To find a pediatrician in Irvine, CA, you can use the following resources: Online directories such as ZocDoc, Healthgrades, and WebMD, Your health insurance company's website or customer service line,....”
On the other hand, ChatGPT could be used to answer general healthcare questions such as “What is deductible in healthcare” or “What is a co-pay.”

ChatGPT could be used to augment a specialized healthcare navigation chatbot, by possibly handling general knowledge questions that the specialized chatbot cannot answer.

Verdict: Can complement existing chatbots to make them more responsive and cover more questions.

3. Health coaching and wellness

These chatbots try to engage patients to achieve a more healthy lifestyle. For example, they may remind patients to exercise and eat healthy every day. Examples of such chatbots include Doppel and 1-million-strong-to-prevent-diabetes. They may also assess the progress of a patient using existing clinical scales, such as PHQ9, which is a popular 9-question scale to assess depression. 

There is no easy way to program ChatGPT to give wellness advice or to monitor a user’s progress over time. Furthermore, ChatGPT does not support administering questionnaires. 

ChatGPT could be used to complement such bots. For example, one could ask ChatGPT if eating pasta is healthy or to suggest ways to burn calories.

Verdict: Can complement existing chatbots to make them more responsive and cover more questions.

4. Healthcare marketing 

These bots typically collect information from leads (potential customers/patients), provide them information on a practice’s services or make an appointment. Such bots are more popular in out-of-pocket specialties, such as dentists, chiropractors, plastic surgeons or physical therapy, which are the specialties that invest more in marketing.

A key intuition here is that as patients search for services on Google, if a practice’s website has a chatbot that can engage the patient 24x7, it will make it more likely to attract this patient, especially in the case of millennials, where instant gratification is key.

ChatGPT would not be very useful here, as it is hard to program it to have a goal (e.g. collect a user’s email) and solicit information from users.

Verdict: Not applicable

In summary, we see that ChatGPT, in its present form, is not applicable for some types of healthcare AI applications, and could be used to complement existing technologies in some other applications. As ChatGPT evolves and becomes more customizable, its applicability in healthcare could also increase.

by Vagelis H.

You have decided it is time to embrace the AI revolution by building and deploying an AI chatbot for your business. There are many options in terms of function, content, live chat availability or AI capabilities. One of these options is where to deploy the bot. The two most popular options are on a Web site or via SMS (texting). We will discuss how to pick the right medium.

Web-based chatbot (the familiar chat icon on the bottom right corner of websites)

Advantages:

• Sleek User Interface: depending on the Web bot provider you pick, the Web bot may have some fancy UI capabilities ranging from buttons to tap, to floating rendering, to auto-complete, to alerts, and so on.

• No need to install mobile app

• Security: most Web bot providers use encryption (HTTPS) to transmit user messages. Some of them even offer domain-specific security and privacy standards such as HIPAA-compliance for healthcare, or CCPA/CPRA or GDPR.

Disadvantages:

• Chatbot cannot initiate a conversation. For example, the chatbot cannot send a reminder the next day.

• Hard to automatically authenticate users (may ask for username/password). This creates friction, as the bot thinks that every time the user visits the Web page, they are a new user. Cookies can alleviate this problem, but they create privacy concerns and also do not carry across devices or browsers.

• Spam possible: users can chat anonymously, so some conversations may be spam. 

These properties make Web bots appropriate for front-desk tasks, or to help a user navigate a Website. They are also good for customer service and live chat escalation.

SMS chatbot

Advantages:

• Great for authentication, as a phone number generally uniquely identifies a user. That is, there is no need for usernames and passwords.

• No need to install an app.

• Chatbot can initiate conversation. Ideal for automated reminders. follow-ups, marketing campaigns, and so on.

• No spam

Disadvantages:

• No fancy user interface, like buttons tapping or autocomplete. Images and videos may display fine, depending on the specific phone model and texting application.

• Weak security and privacy. For example, it is not natively HIPAA-compliant, but could be used to exchange sensitive information with the user’s consent (acknowledging the risk).

• More expensive, as the mobile carriers may add a fee, on top of the cost of the chatbot platform (or the chatbot platform has to increase its per passage fee to cover the carrier fee).

These properties make SMS chatbots ideal for automated long-term engagement of users, or marketing campaigns (one has to be careful not to violate anti-spam regulations). 

Specifically, mass texting or drop campaigns can be built using SMS chatbots. For higher throughput, short code texting can be used, by purchasing a short code, which supports higher rate of messages per second. These bulk text campaigns have been shown to have much higher open rates than email campaigns.

SMS chatbots can also be useful for user-initiated conversations (as is the case for Web bots). For example, the user can text a keyword like “help” to a phone number to start chatting. 

For completeness, we also briefly discuss other possible media to deploy chatbots.

Facebook Messenger/Whatsapp/Viber

Advantages:

• Can authenticate user based on their username or phone number

• Good for international audience.

• No need to install app (assuming user already has Messenger or Whatsapp app installed)

• Can push notifications and follow-up messages to users

Disadvantages:

• Not as customizable as the web bots.

• Strict control and regulation by Facebook. Require a lot of time to keep these chatbots compliant with the continuously changing policies of Facebook.

• Not private, as there is an entity in the middle that monitors the traffic, although some claim end-to-end encryption. Not HIPAA-compliant either.

Voice (Alexa, Google, Siri, Cortana)

Advantages:

• Voice; can access while driving or cooking or other activities that require the user’s hands

Disadvantages:

• Hard to authenticate, as often the whole family is sharing one account or device

• Cannot exchange files or images

• Some claim HIPAA compliance

Mobile app

Advantages:

• Full control on the user interface. Could also support voice, although it is not common

• Support mobile notifications and follow messages

Disadvantages:

• User has to download app (recall that chatbots became popular as a replacement of apps)

• Higher development cost, as mobile app has to be created and maintained

• High-level of security and privacy possible

Part of a larger Web application (such as a Patient portal in healthcare)

Advantages:

• Supports central user authentication and integration with other user data

• User has to login to larger Web app

Disadvantages:

• High development cost, as chatbot has to be developed as part of a larger project

• High-level of security and privacy possible

by Vagelis H.

Building effective chatbots is challenging for several reasons. 

Chatbots must first understand the user’s information need, as expressed by one or more dialog utterances. Then, the chatbot must figure out how to best respond, by combining available information which may be in the form or flow diagrams, FAQs or documents. An often ignored attribute of chatbots is that they must show empathy and emotionally connect with the user, to maximize their effectiveness. 

A recent article by Prof. Rosalind Picard of MIT identifies three key principles that chatbots must possess to build great relationships with humans, which are the same principles that a human must possess to build great relationships with other humans. Here, we list these principles and discuss how these can be applied to chatbots (the original article mainly focuses on human-to–human relationships).

Principle 1: Continuously collect feedback from users and use this feedback to improve the chatbot.  For example, at the end of each conversation the chatbot may invite the user to take a short survey. Or right after a response that the chatbot thinks that it may be inaccurate, ask the user if this response is what they expected and if not, why. 

Principle 2: Make sure the chatbot correctly understands the user’s information need. For example, if the user says “how much is it?” the bot can reply “are you inquiring about the price of the Botox procedure?” That is, the bot could repeat the user’s question, paraphrasing it and adding missing context (possibly collected from previous turns) to confirm that it properly understands the user’s need.

Principle 3: Try to generate active, constructive responses. Such responses make people happy and help build stronger relationships. For example, if the user shares some good news, respond enthusiastically, instead of just saying “congratulations.” 

by Vagelis H.

Marketers from all industries have always relied on outbound campaigns to reach potential customers. Before the time of the Internet, letters and phone calls were the main means to reach customers.

Email Campaigns

Then email blast campaigns became popular, as they offer a cheaper and more efficient way to reach out to customers. email campaigns evolved into email drip campaigns, where a marketer can design a sequence of emails that will be sent over time to leads. Conditional drip campaigns are also possible, where the next email depends on whether the lead opened a previous email.

Email campaigns suffer from low open rates. Further, they are unable to effectively complete a conversion. For example, the email may encourage a patient to call the doctor to schedule an appointment or send a customer to a website to shop for concert tickets.

The key players in email marketing are Mailchimp, Sendgrid, Active Campaign, and many more.

Texting Campaigns

After email campaigns, came texting campaigns. The key promise compared to email campaigns is higher open rates. A marketer needs to have a list of phone numbers instead of a list of emails. The message of course has to be shorter, typically an image plus a small paragraph of text, possibly with emoticons, and a hyperlink, to be sent as one or more MMS messages.

Similar to email campaigns, texting campaigns must have a way for a user to opt-out, e.g. type “STOP”.

Texting campaigns indeed increase the open rate, however they still suffer from the second drawback of email campaigns: they cannot complete a conversion. Instead, they typically send the lead to a website to complete the transaction. This incurs unnecessary friction and lack of instant gratification. For example, the text message may encourage a user to make an appointment, but what if the user has a question on insurance coverage or cost?

Similar to email drip campaigns, one can create, with some of the tools out there, drip texting campaigns, where a lead receives a sequence of messages over a period of days or weeks.

There are several products that enable texting campaigns, such as EZTexting and Simple Texting.

Chatbot campaigns

Chatbot campaigns are the evolution of texting campaigns, where a single message is released by a chatbot. That is, instead of blast sending a message to thousands of leads, we blast trigger a chatbot to the leads. For example, a texting campaign could send out message “It is time for your dental cleaning. To schedule your cleaning, go to https://mydentist.com/cleaning”. Instead, a chatbot campaign could trigger a chatbot that initially sends out message “Hi Bob, it’s been a while since your last cleaning. Is now a good time to chat about it? Yes/No.” If the user replies “No” the chatbot will try again the next day, else the chatbot will continue with message “Great, let’s get you scheduled. Has your insurance coverage changed since your last visit? Yes/No”, and the chatbot can then continue to collect all needed information and find a suitable time slot. This increases the conversion rates, improves user experience and decreases customer service cost (saves calling center time).

SmartBot360 is a leader in chatbot campaigns. It allows users to easily visually create chatbot flows and then create conditional drip campaigns around these chatbots. To read more details about how SmartBot360 supports texting campaigns, read the relevant documentation page.

An added benefit of SmartBot360 is that it supports HIPAA-compliance so it is possible to deploy chatbot campaigns in privacy-critical industries such as healthcare.

by Anthony Y. updated 10/29/2021

Whether you own a small healthcare office or manage the customer experience in a large healthcare company, there are many ways to use an AI chatbot to accomplish business goals. For small offices, a chatbot is usually used to provide support 24/7 by responding to most questions while routing more complex situations to a staff member, but typically chatbots built for small offices do not need to be too complex and can be easily set up using one of our many healthcare templates.

Large healthcare companies have different goals and needs, but chatbots are typically used as an additional way for patients to get instant support or as an interactive patient portal, ultimately reducing call center load by utilizing conversational AI to respond and guide patients. There are many ways to customize SmartBot360’s chatbot, but due to compliance and complexity of integrations, we recommend deploying a simple FAQ chatbot then adding more integrations over time as other chatbot use cases become apparent. Whether a company wants to integrate member ID login into their chatbot, schedule automatic follow-ups with patients, or handle screening prior to appointments, our healthcare chatbot experts can work with you to customize a chatbot to be as complex as needed.

Use of chatbots in healthcare

PATIENT TRIAGE & NAVIGATION

Help patients quickly navigate the complexities of the healthcare system using an AI chatbot with natural language understanding. SmartBot360 integrates with your workflow and guides each patient according to triage guidelines. Keep logs of each patient neatly sorted and referenceable whether the patient needs digital or in-person help with our CRM integrations or through SmartBot360’s backend API. 

Example

FAQ Chatbot

SmartBot360’s chatbot can be set up to answer common questions. Whether a patient types a question in chat or follows a setup flow, SmartBot360’s AI understands the context and responds accordingly.

REVIEW FILTERING

Getting reviews helps with improving visibility and conversion rates. When you schedule a chatbot to automatically message someone a few days after their visit, there is a higher chance of them leaving a review whether they are satisfied or unsatisfied. Use our chatbot to automatically message patients to leave a review a few days after their visit. SmartBot360’s AI allows for filtering of reviews by sending patients with positive experiences to a Google or Yelp page, while sending a link to message customer support to get further support for their problem instead of having them leave a negative review.

TESTING & VACCINE INFO

Relieve support staff with a chatbot flow to automatically provide answers for patients looking for testing & vaccine info. Reduce call center load and free up time by answering repetitive questions and guiding/setting up appointments automatically.

Example

AUTOMATE CUSTOMER SERVICE

Many patients ask repetitive questions that take up valuable front office staff time that can be answered directly by our chatbot.  Think of an interactive FAQ chatbot where patients type their questions and get an immediate answer without waiting on hold or for an email reply from your staff.  The chatbot improves clinical workflows and both patient and staff satisfaction. 

HIPAA-COMPLIANT SCREENING & QUIZZES

Screen patients questions prior to their appointment or ask website visitors a series of questions to then recommend a service. We’ve noticed the more a person interacts with the chatbot, the more they will recognize and trust your business. Using quizzes increases the chance of scheduling an appointment or coming back to your website after exploring other options because of the more “personalized” recommendation.

LONG TERM PATIENT ENGAGEMENT & FOLLOW UP

Set up a chatbot to automatically follow up or message a patient when it is time for their next checkup or check-in a certain amount of time after their procedure to make sure they are following proper steps to recover.  Effective follow-up leads to repeat business and more positive reviews.

by Vagelis H.

In a previous article, we discussed how a chatbot can leverage a document collection to respond to user questions. The simpler method is to return a document snippet that is relevant to the question, while a harder method is Question-Answering (QA), which tries to generate the actual answer to the question.

SmartBot360 is at the forefront of such document-driven chatbot capabilities, which enable the creation of effective customer service chatbots. In this article, we discuss the ways that SmartBot360 offers to enable document-driven AI in a chatbot. 

For a live demo, check out https://smartbot360.com/demo/docai, which demonstrates the various ways to query documents in SmartBot360.

The first two ways return a relevant snippet, while the third attempts to directly answer the question.

1) Document Search --DEPRECATED AS OF 5/27/24-: This method allows a chatbot to search the pages of one or more web sites. The bot admin first specifies a set of URLs (e.g. smartbot360.com/documentation) to be indexed, and then inserts a Document Search box in the flow diagram of a bot. The bot then asks the user to input a question and returns the most relevant snippets, along with links to the documents that contain them, as shown below.

2) Google Programmable Search integration --DEPRECATED AS OF 5/27/24-: This method is similar to the first one, except that here the documents are indexed by Google and not by SmartBot360. Specifically, the admin must first create a Google Programmable Search Engine, and then link it to the SmartBot360 account. Then, the admin inserts a Google Programmable Search box in the flow diagrams, which works similar to the Document Search box. A disadvantage of this method is that it may not be HIPAA-compliant, because Google Programmable Search Engine is not HIPAA-compliant.

3) Question Answering --DEPRECATED AS OF 5/27/24-: This is the most advanced method, as the actual answer is searched, using AI. The documents are indexed in the same way as in Document Search. Then, a QA box is inserted to the bot flow diagram. The bot then asks the user to input a question, and it returns the best answer to it, along with the link to the document where this answer was found, as shown below.

More and more healthcare businesses are taking advantage of digital services, but with so many new technologies out there, it is best to keep things simple while providing access to as many patients as possible. With 96% of text messages being read within 3 minutes, utilizing HIPAA-Compliant texting can minimize no-show appointments, improve engagement between provider and patients, and even provide more effective post-procedure instructions—all while saving your practice time.

Is Text HIPAA-Compliant

As SMS is handled by native apps on most phones, it is generally not going to be HIPAA-Compliant. Scheduling, general information, and non-sensitive information can be sent through text, but when Protected Health Information (PHI) needs to be sent, SMS is not considered HIPAA-Compliant.

Aside from using SMS for reminders and general business information, it is generally safer to prohibit texting because of the potential security risks. There are uncontrollable risks of patients accidentally providing PHI through SMS, losing their phone, and more, but because they may not have knowledge regarding HIPAA compliance, what happens after these scenarios may cause problems. 

The other elements of HIPAA compliance, like the need for encryption, access, and audit controls, do not exist in native SMS apps. HIPAA necessitates that accessing, creating, modifying, sharing, or deleting PHI needs to be recorded, enforcing audit controls for texting is not possible yet because it is not a priority for SMS developers. Service carriers can technically also intercept messages since they are not encrypted, but overall there are many vulnerabilities associated with SMS.

This is why with SmartBot360, we focus on HIPAA compliance and adapted our chatbot to work with SMS apps as well. SmartBot360’s AI detects potential compliance issues, or if the chatbot is asking a question that may require the patient to enter PHI, it sends a secure link that requests them to continue the conversation and input their health information there.

Things to do with HIPAA-Compliant texting

While it is standard to follow-up with calls or email, SMS has mostly been the channel that gets the most impressions even when they do not reply. Everyone at least sees a text on their phone, whereas someone not picking up a call or going over their email may miss it.

Because SMS recipients see the general gist of the message before swiping to delete the notification, reminders for appointments are rather effective as patients can be notified a day before their appointment, and if something new came up, they can simply reply to cancel it so that the provider can save time and fill in the spot with another appointment. 

Providers can also schedule chatbots to follow up with patients after their visit to request a review, and with SmartBot360’s chatbot SMS follow-up, it can route positive reviews to the Yelp/Google page and route negative reviews directly to staff so that issues can be directly resolved without having a patient leave a negative review.

Pre-visit planning can also be handled with HIPAA-compliant texting. Prior to their in person visit, the chatbot can be scheduled to send pre-visit planning forms so that the patient can fill it out and save more time during the actual check-up instead of filling out forms at the office. It also provides a quicker way for doctors to glance over all their information so they can dive deeper into the potential symptoms and solutions.

There are many ways to use HIPAA-Compliant texting for your organization, and with SmartBot360, it can be customized to handle simple tasks or complex tasks depending on your organization’s texting goals.

by Vagelis H. 04/29/2021

Provider time is a scarce resource. AI technology is well-positioned to automate some of the components of the provider-patient interaction, to make visits more efficient for all parties. Chatbots can be effective at collecting patient information before a visit takes place. Two of the most critical operations where chatbots can play a significant time-saving role are:

• Social Determinants of Health (SDoH) screening. SDoH can account for up to 80 percent of individual health outcomes.1 Periodically collecting SDoH information can improve the quality of care, save many minutes of the provider’s time during a visit, and refer to the right community services in a timely manner. 

• Pre-visit planning is a proactive approach to care that ensures each visit is meaningful and well-organized by shifting administrative tasks out of a patient’s appointment time so team members can focus on the patient — not paperwork.2 With pre-visit planning, providers have the information they need at the time of the visit, and the care team can address the patient’s needed services based on evidence-based care. Plus, staff have more time to collect copays, schedule necessary follow-ups and perform other tasks that must be completed while the patient is onsite.

For chatbots to be effective in these (or other similar) tasks, they must satisfy several requirements:

1. Be HIPAA-compliant. For example, collecting this information over SMS or Facebook Messenger is not HIPAA-compliant. However, these non-HIPAA media can be used to initiate a chat and then move the conversation to a secure medium by sending a hyperlink.

2. Be frictionless (see our previous article about frictionless chatbots). Asking patients to remember a username and password, and login to a patient portal (like MyChart) to enter this information is an example of a high-friction solution. Asking them to download and install a mobile app is even higher friction. An effective solution is to send an automated SMS message to the patient (thus using the phone number as an authentication mechanism), which includes a secure hyperlink (for HIPAA-compliance as mentioned above). 

3. Integrate with the patient’s health record. Depending on the size or nature of the organization, this means integrating with a large EHR system like Epic (through an integration protocol like FHIR), or a smaller one like DrChrono. In addition to EHRs, integration with an organization’s workflow may be achieved by sending notifications to the right staff members (e.g. case managers) when action is required.

SmartBot360 provides a platform and expert healthcare consultants to easily design automated SMS-based HIPAA-compliant chatbot workflows to achieve efficient SDoH screening and pre-visit planning.

References

1 Sanne Magnan. “Social Determinants of Health 101 for Health Care: Five Plus Five” National Academy of Medicine https://nam.edu/social-determinants-of-health-101-for-health-care-five-plus-five/

2 https://edhub.ama-assn.org/steps-forward/module/2702514

by Vagelis H. 03/29/2021

Chatbot platforms, such as SmartBot360, generally generate a short HTML snippet that you copy-paste to your Web page and this is it, right?

Not really. There are several options to consider that can have a big impact on the degree of engagement, that is, how many page visitors interact with the chatbot and for how long. We analyzed statistics across healthcare web sites that host chatbots and discuss the results.

The choice with the biggest impact is if the Web bot is open when loading a page, or if the patient has to click on the bot icon to open its window. Below we show the three basic options:

(a) The Web bot is open on page load

(b) The Web bot is closed on page load, and has a “Click to chat” or other similar popup message to make it more visible and make it clear what the bot icon is

(c) The Web bot is closed on page load, and has no popup message, which makes the visitor much less likely to interact with it.

Option (a)

Option (b)

Option (c)

The bot engagement rate is the number of patients that interact (send at least one message) with the bot divided by the number of people that visit the page.

For option (a), the engagement rate is around 25%. For option (b) it drops to about 11%, and for option (c) to around 5%.

When the bot starts closed (options (b) and (c)), among the patients that click the bot icon, 74% of them continue and chat with the bot, whereas the rest 26% do not send any message.

A subcategory of option (c) is when the bot starts by clicking a CTA button that is often prominently displayed. In this case, the engagement rate is around 15%, depending on the position and size of the button.

These percentages are based on chatbots with frictionless flow, that is, chatbots that do not start with a personal question (such as what is your name or email) or with a disclaimer (such as “I am a chatbot, so I cannot offer any advice, click to proceed”). If such higher friction messages are the first messages in a chatbot, the above percentages drop by about 40% across the board.

In terms of length of a chat session, as expected, when users have to click on the chat icon (options (b) and (c)) they are more likely to have longer sessions. We found that these sessions are about 20% longer (in number of messages) on average. 

Another option that is sometimes used when displaying a web bot is to embed it on the web page as shown below. The effect of this option depends on the position of the embedded bot within the web page.

To conclude, we have found wide variance in the bot engagement rates depending on how the web bot is displayed. We always advise SmartBot360 customers to try out different settings, to pick what works best for them.

by Vagelis H. 01/26/2021

We live in the time of AI and expectations are high. I had several customers ask me about the extent of AI used by SmartBot360.  A common question is if a chatbot can become smarter as more people chat with it, which I cover in an earlier article. Another question, which is the focus of this article, is if a chatbot can search for an answer in a Web site or other document collection (such as SalesForce documents).

The answer is that there are several techniques to achieve this, with different levels of sophistication, which lead to different user experiences. Below, we will examine some of these solutions.

1) Search for relevant documents

This is the simplest solution. The key idea is that if the user asks for something that the chatbot cannot answer, for example, “can I bring my dog to the visit,” the chatbot may search the documents of a web site and send the user to a relevant document, like sending the user to the page visit_policies.html. 

The chatbot may recommend only the top-1 most relevant page or a small list of results. The ranking can be simple keyword matching or use AI to search semantically for the most relevant document. The most recent techniques to achieve accurate text matching are based on deep learning, and specifically on computing the similarity between document-to-vector embeddings.

Another consideration is how to incorporate the context of the chat to the query. For example, if the user mentioned before that she is interested in the Riverside office, then the query should search if the Riverside office allows dogs. Most solutions out there do not support context.

There are a few chatbots that employ this document search approach. For example, Anthem has a simple keyword match chatbot in its members portal. Typing “find doctor” forwards the user to the provider search page. SmartBot360 includes a Document Search box in its bot diagrams, which returns the top-5 most relevant documents given a query, as shown below. SmartBot360 supports several document sources, including Web page. SmartBot360 also has options to improve the search accuracy by using the conversation context.

2) Search for the answer segment

The problem with the first solution is that it returns whole pages rather than the answer to the user question. Going back to the example of the user asking “can I bring my dog to the visit,” if page visit_policies.html contains a sentence “For safety reasons, no pets are allowed to our offices.” then the chatbot could return this sentence instead of the whole page.

This is much more challenging than the first solution, but fortunately, there are recent advances in deep learning that bring this solution closer to reality. There is a recent competition set up by Stanford to promote research in this area. SmartBot360 is testing several solutions to extract the best text segment by customizing state-of-the-art techniques.

3) Give the answer

The ultimate goal is to answer the question as a human would. Going back to the example of the user asking “can I bring my dog to the visit,” the answer should be “No, you cannot.” 

This is called Question-Answering (QA) in the research literature. QA is the holy grail of search (including Web search), and even though many researchers and companies have been studying solutions, most work only for limited types of questions. For example, take a look at Google AI’s research page on this.

by Anthony Y. 01/04/2021

Are you a healthcare professional looking for an easier way to engage with your patients? You may be interested in our AI healthcare chatbot – it’s one of the few chatbots available that makes use of HIPAA-compliant chats. SmartBot360 is an AI-powered healthcare chatbot that can be used in any setting. Whether you’re a local physician or you’re running a national healthcare chain, SmartBot360 can be of assistance. It can be used to organize patient interactions on a small or large scale. While the chatbot can be set up in minutes for a basic workflow, it can also be customized infinitely to match the demands of your customer base. 

How Can SmartBot360 Help Me Manage More Patients More Effectively? 

You may be wondering how a chatbot can directly influence your patient management in an effective manner. Taking just minutes to set up with no coding necessary, here are just a few of SmartBot360’s key features that will allow your healthcare office or website to attract more patients. 

Optimize Patient Screening and Increase Patient Engagement 

SmartBot360 can be used for optimized patient screening and engagement, so much so that enterprises see reduced call center load and small offices may find themselves with an extra (or full-time) receptionist on their hands. SmartBot360’s HIPAA-compliant nature gives you everything you need to automate your general patient screening procedure. You’ll notice reduced drop-out and 20% higher patient conversion rates when using our platform to optimize SmartBot360’s HIPAA-compliant chat flow, and our flow-diagram chatbot builder allows you to easily change the way your chatbot interacts with patients – no coding required. 

If you’re working in healthcare on a larger scale, you can customize your SmartBot360 to serve as an improved customer service chatbot. SmartBot360’s HIPAA-compliant chat flow will enable your healthcare enterprise to legally, safely, and thoroughly handle patient questions, appointment notifications, and even online wellness checkups. SmartBot360’s integrative properties with popular healthcare platforms like Acuity Scheduling and DocuSign ensure a seamless transition between commonly known healthcare applications. More integrations can be added to your SmartBot360 upon request. 

Upgrade Your Customer Service and Patient Wellness Capabilities with Ease 

SmartBot360’s HIPAA-compliant nature would be remiss if it didn’t improve your customer service capabilities drastically. However, SmartBot360 does that and more to assist your healthcare establishment with improving your complete customer life cycle. 

SmartBot360 will engage and guide visitors on your website effectively. It can answer real visitor questions and provide immediate assistance to visitors who may be suffering with medical maladies. SmartBot360 can recommend products and services to visitors while remaining HIPAA-compliant. In providing this immediate, HIPAA-compliant assistance, SmartBot360 improves your patient conversion rate by up to 20%. 

SmartBot360 can also deliver superior patient care on a consistent basis by obtaining relevant medical information from your patients. This allows your healthcare establishment to more effectively conduct patient assessments and provide related medical information or assistance, and in some cases will allow your establishment to redirect patients to appropriate care practitioners. 

By providing instant, 24/7 support to your visitors, your healthcare establishment can eliminate wait times (a major factor that causes visitors to your website to seek medical assistance elsewhere). Because SmartBot360’s HIPAA-compliant nature uses conversational language and can understand patient responses, it can provide medical recommendations and answers to your website visitors instantaneously. This increases the likelihood that your establishment will convert visitors into patients. 

Promoting your patients’ well-being and health has never been easier with the help of SmartBot360. By providing proactive chat notifications and becoming invaluable in the patient treatment process, SmartBot360 is able to help your patients stay on track and see results on a consistent basis. 

If you’re interested in multiplying your patient conversion rate, screening and qualifying leads, improving your customer service baseline, making more appointments with more patients, and combining SmartBot360’s HIPAA-compliant flow with live chat while meeting privacy and security requirements, give SmartBot360 a try today! 

by Vagelis H. 12/31/2020

This article focuses on healthcare chatbots, so we will discuss how to make a HIPAA-compliant chatbot be GDPR-compliant as well. This is necessary if a healthcare entity plans to interact with EU patients, in addition to US patients (HIPAA is a US regulation).

For a discussion on how to make a bot HIPAA-compliant see our previous article.

Similar to HIPAA, GDRP is a set of guidelines and not technical directives, so ensuring GDPR-compliance is a joint responsibility of the chatbot designer and the chatbot platform. For example, the chatbot must request permission to store and use data provided by the user; this is generally the responsibility of the chatbot designer. On the other hand, securing the data using encryption or other techniques is a primary responsibility of the platform.

Here are the key additional measures you have to take to make your HIPAA-compliant chatbot also GDPR-compliant:

1. (platform responsibility) Report data breaches within 72 hours, whereas HIPAA requires reporting within 60 days or more.

2. (platform responsibility) Right to erasure (also known as “the right to be forgotten”), that is, have a mechanism in place so users may request that all their private data is deleted. You must delete the data within one month from the time the request is received.

3. (bot designer responsibility) Explicit consent for storing and purpose of using data outside direct patient care. For example, the bot could start with a message “Do you agree to have your responses stored and used to improve your user experience? Yes/No”

A link to a terms page is generally not adequate.

This is similar to the ubiquitous cookies consent messages we see in almost all Web sites recently, to comply with GDPR.

The platform could force adding a predefined message at the beginning of each chatbot, but the bot designer is in a better position to make the message more meaningful and insert it in the right position, given the purpose of a chatbot. Also, some chatbots may not need consent, as they don’t collect any personal information, for example, anonymous FAQ bots.

4. (joint responsibility) Consent for transferring data outside the EU. The simplest way to handle this would be to store the chatbot data in the EU, for example, in EU data centers of AWS. If you prefer to store the data outside the EU, you need to get consent from users from that. You could add this to the consent for the purpose of the chatbot (Measure 3)

Business agreement: Similarly to how companies must sign a Business Associate Agreement (BAA) for HIPAA, for GDPR there is an equivalent Data Processing Agreement (see template).

Chatbot media: As discussed in a previous article, only Web-based chatbots are HIPAA-compliant, due to the third party in-the-middle complication. This restriction is not necessarily true for GDPR. For example, there can be a GDPR compliant bot on Messenger. However, the company (Data Controller) and not the chatbot platform (Data Processor) would typically be responsible to make sure that Messenger deletes the user data when requested (right to erasure).

by Vagelis H. 12/15/2020

Anybody who has worked on creating chatbots for healthcare providers (or any other goal-oriented chatbot) soon realized that there is no magic AI. In commercial chatbot platforms, AI is used in limited aspects of the chatbot, such as intent classification (e.g. patient has billing question vs. tries to make a new appointment) or information extraction (e.g. map “tomorrow” to a date). For more details on AI and its limitations for healthcare chatbots, read our previous articles on bot capabilities and bot technologies.

In all commercial platforms, the chatbot creator has to manually define some type of flow diagram (or slot filling), which must try to anticipate as many as possible inputs from the patient, while advancing the conversation towards the chatbot’s goal (e.g., make an appointment).

SmartBot360 has been working with a large number of healthcare providers and organizations since 2018, building and refining chatbots to attract new patients or serve existing ones. Here are some lessons learned, which can help in creating more effective chatbot flows:

1. Allow free text input only if you have free text processing. For example, if your chatbot can only understand “yes” or “no” for a specific question (e.g. “are you a new patient”), and any other input will lead to “sorry, I didn’t get it,” don’t allow the patient to input anything else. This could be achieved by selectively disabling the free text message box during some parts of the bot.

2. The most common “uninvited” question is “do you accept insurance X?” If you are a healthcare provider, no matter what the chatbot’s welcome question is, e.g., “what’s your name,” many patients will immediately ask about insurance. The reason is that they don’t want to waste their time interacting with the bot and then find out they are not covered. So, every healthcare chatbot must have a mechanism to catch such questions. The simplest way is to have a text-contains component that matches terms like “insurance”, “blue cross”, “anthem” and so on.

3. Appointment (re)schedule. Close second after the insurance-related questions are appointment rescheduling questions e.g., “I want to reschedule the appointment I have tomorrow morning.”

4. List of services. Another common category is questions about the offered services, e.g., “do you do liposuction?”

5. Bot or human? Another popular question in healthcare bots, as in all bots, is something like “are you human or bot?”

6. Live chat. A large number of patients may also ask to speak to a human at any point.

A key challenge to address all the above cases is that these patient utterances can occur at any point during the chatbot flow, which we refer to as “off-script” utterances. Some chatbot platforms may allow designers to define different intents for each of these patient requests. A problem with that is that it breaks the flow of a goal-oriented chatbot. Another approach, employed by SmartBot360, is to augment flow diagrams with a knowledge base of common questions and a language generation module.

by Vagelis H. 12/02/2020

As in-person contact has become increasingly hard with all the quarantine measures, patients turn to online options to get information on COVID-19, evaluate their symptoms or get advice. Live chat and telemedicine services, such as CVS Micro Clinic Video Visit, are overloaded and have long waiting lines.

Hence, there is a need to turn to more scalable solutions, and chatbots are such a solution, as they are good at scaling routine chat interactions. So how can chatbots be used?

Use Case 1: Answer common questions

A simple and effective use is to load to the chatbot FAQs from reliable sources like CDC, WHO, or local healthcare agencies. SmartBot360 has created such a COVID-19 FAQ bot, which is made available to its customers (or free trial accounts) as a free chatbot template. Similarly, WhatApp has built a bot for common COVID-19 questions. Once an initial FAQ chatbot is built, an administrator should maintain it, by updating the information and adding new content.

Use Case 2: Screen patients

A chatbot can ask questions to the patient and the probability that the patient is infected with COVID-19. The questions are typically a combination of demographics (age, gender, location), symptoms (shortness of breath, fever) and lifestyle (recent trips, daily activities). As an example, CDC has partnered with Microsoft to create such a screening bot. As another example, Jiseki Health is using SmartBot360’s platform to deploy state-specific screening chatbots on SMS.

Use Case 3: Push customized information to the right population segments

This is the natural next step after screening patients for COVID-19. For example, if the screening showed that the patient is high-risk because of her lifestyle (e.g. works at supermarket), the chatbot could send daily reminders and tips on how to protect herself or how to get tested close to her work. As another example, if someone lives with elderly at home, the bot should remind them of the risks of his lifestyle to them. Such follow-up bots must be deployed on non-Web media like SMS, Messenger or WhatsApp, as it is not possible for a Web bot to initiate a conversation.

This last use case is still to be realized to the best of our knowledge. Let’s all try to do our part using the best available tools to mitigate this health crisis.

by Vagelis H. 11/13/2020

A key promise of chatbots is their potential to increase patient engagement. This in turn can lead to better health outcomes, lower costs and more successful patient acquisition. Before chatbots became popular, health organization relied on mobile apps to engage patients. However, this never succeeded because of the high communication friction. Specifically, the patient has to first find and install the app, then create an account, a login and a password, and remember them.

So, how can chatbots reduce this friction? The first part (installing the app) is the easy one to disrupt, as a chatbot can be hosted on a Web site, Facebook Messenger, SMS or Whatsapp, just to mention a few, so there is nothing to install. The second part is more challenging.

Bots that are hosted on chat platforms (like Messenger, SMS or Whatsapp) are not HIPAA-compliant. Web chatbots can be HIPAA-compliant, but without a registration (username/password) process it is hard to authenticate (i.e. ensure that the individual is who he or she claims to be) a patient.

This means that Web chatbots can be easily used to collect new leads, but not to engage the same patient in the long term. When the same patient uses a Web chatbot on two different days, it is generally not possible to know that this is the same patient, as the IP address or the browser’s User Agent are not reliable authentication methods.

So, how can a chatbot authenticate a patient without requiring installing an app or using a username and password? One way, which is being employed by the SmartBot360 platform, is to authenticate the patient via their phone number and then switch (by sending a unique hyperlink) securely to a HIPAA-compliant Web bot. There are some details to work out, for example, ask permission from the user to contact them via SMS and also to notify them about health-related matters.

Achieving frictionless patient experience is clearly a great challenge, so we may see some promising technologies in the future that make it easier. For example, mobile phone browsers in the future may have protocols to authenticate Web chatbots. Browser notifications are a step in this direction, but they are not enough as they do not authenticate the user.

by Anthony Y. 11/11/2020

Chatbots are one of the more subtle technological marvels to have emerged from the digital revolution of the past two decades, and yet they are one of the more omnipresent ones. And with how good they’ve gotten with the exponential leaps in artificial intelligence and machine learning in the past few years, chances are that most websites you could think of are already using a chatbot in some way or form without you even knowing it!

Websites designed to cater to the healthcare sector are no different, and most websites of hospitals, clinics, insurance companies etc. have a sophisticated chatbot that takes care of tasks that can be automated easily.

And while the chatbots in healthcare vary from simple reminder-based ones that can suggest diagnoses like Your.MD or Florence to complex HIPAA compliant all-encompassing ones like SmartBot360, for the purpose of today’s article we are focusing on the latter kind.
The HIPAA compliance is a major factor, for those bots can not only register the information that a user enters but also share them with the medical service in question. From a process flow perspective this ends up being a godsend as far as the customer experience is concerned. The sheer degree of engagement a chatbot enables in something as simple as scheduling a visit to the general practitioner augments the customer experience in a way that sweetens the whole process!

Chatbots and Better Quality of Healthcare

The implicit benefits of chatbots are as great as their explicit ones. The most obvious of these implicit benefits is how chatbots free up the human resource involved in the healthcare value delivery chain. With repetitive and time-consuming tasks such as information retrieval and appointment scheduling now handled by a few lines of code rather than dedicated personnel, said personnel can devote their energy and time towards patient servicing. This move goes a long way towards improving both the efficiency as well as the quality of the healthcare being delivered to the patients.

Continuing this theme of efficiency, chatbots are also not beholden to the constraints of on call hours the way medical personnel are. Chatbots are available to users 24 hours a day, 7 days a week. The best way to demonstrate the benefit of this is through a mini case study:
Imagine it’s the middle of the night and you’re feeling what feels like an oncoming bout of food poisoning. Now the symptoms do seem to point in that direction, but you can’t ever be too sure unless you’re a medical professional, can you? Speaking of which, since it’s the middle of the night you can neither visit one nor even call one. Waiting until the morning seems difficult, but wait, you can ask a chatbot! By merely entering your details and elucidating upon your symptoms, you can have the chatbot index that information against its data repository and offer you a reasonably accurate diagnosis to proceed further with instantly. And should you feel a practitioner’s visit is still called for when the morning comes, you can easily have the chatbot schedule it for you as well without having to wait until the offices open. 

As the above case study illustrates, chatbots allow for a degree of seamlessness to the entire value delivery funnel in healthcare that fundamentally betters the customer experience. The most important way to get it right is to lock down the chatbot’s live chat feature. Since customers are engaging with the chatbot in live conversation the same way they would with a human, more and more chatbots are focused on nailing the tone and cadence of their responses. As artificial intelligence and machine learning make greater advances, there is a beefier framework supporting the chatbot. Most modern chatbots today can engage in live conversation that is a close approximation of human speech patterns, thus making the entire act of engagement feel as smooth and seamless as that of talking to a human.

Improved User Journeys Lead to Improved User Experience

Chatbots are designed to be as minimally intrusive as possible, and often remain nested in a little corner of the page out of the user’s line of sight until they need to reveal themselves. This act of revealing usually happens when the user performs an action that leads to a set of predetermined actions, and the chatbot serves as a guide to walk the user through the rest of the process. To illustrate with another example, let us assume that a user is browsing through the website of a clinic when they click on the timings section. Since it is a reasonable course of action for the user to schedule an appointment next, the chatbot will pop up and offer that service! This seamless and contextual approach to user prompts makes the experience very fluid and engaging and is something that technology offers that a real-world equivalent can’t with similar ease.

And ease does end up being of paramount importance since in the absence of chatbots, the next best alternative to automate tasks with minimal human intervention is the use of interactive voice response (IVR) systems. Once heralded as revolutionary, IVR systems are now widely despised as they are seen as a slow and inefficient halfway house between effective automation and phone calls. With very high drop-out rates amongst a larger and larger proportion of surveyed audiences, chatbots are perceptibly better than IVR systems. The biggest advantage offered by them is the speed – with instantaneous responses in real time, users no longer need to press a number and wait for a good while to hear a pre-recorded message play out in its entirety before pressing another number and having the whole affair repeat ad nauseum!

With the use of chatbots, another folly of IVRs that is addressed is how dry and “annoying” they can be to have as companions through the user journey. While IVRs had to focus on robotic sounding pre-recorded messages with clear pronunciation that would be played through in its entirety, chatbots offer textual exchanges that can be read through instantly.
Moreover, chatbots can be programmed to be an extension of the healthcare partner’s particular “brand”. Thus, chatbots can be anthropomorphized to reflect the brand image the healthcare provider is trying to project, and the chatbot’s responses can accordingly be professional, conversational, funny etc. A feeling of live conversation, fast visibility and engaging interaction makes chatbots a valuable tool in improving a customer’s overall user journey, and thus in turn, the overall user experience itself.

Of course, a tangential benefit of using chatbots ends up being that all the aforementioned can come together to replace boring old forms with more creative and engaging methods of data collection from the users. Whereas previously users would have to fill in the long, exhausting forms at their practitioner’s offices, the same data can be collected by the chatbots in a very engaging and quick manner to be shared and called upon as needed. This also eliminates the complexities that traditional forms would sometimes bring requiring addressing by someone at the practitioner’s office by simply letting the chatbot frame the question more simply relying upon context cues that can be gathered from the responses!

Putting it All Together…

Chatbots are an important tool for any industry but have particular advantages to offer as far as the healthcare sector is considered. With most chatbots today focused on patient engagement, the user experience ratings are stellar and are only going to keep improving as chatbots become more and more advanced to a degree where they can mimic human interaction in a way that makes the customer experience involving a chatbot virtually indistinguishable from that of one involving actual human interaction.

With the world becoming more digital and more isolated, that sounds like a good thing.

by Anthony Y. 11/06/2020

Having a chatbot on your website to engage new website visitors increases conversion rate by giving website visitors an easier way to get answers for their immediate questions. It can simply be used as a standalone chatbot without any additional tools or integrations, but it is useful to take advantage of all the ways to maximize the usefulness of the chatbot.

See SmartBot360’s Integrations Here

Facebook Tracking And Retargeting

Chatbots usually have a way to insert custom code or to simply insert the facebook pixel ID for Facebook tracking. In SmartBot360’s case, you simply need to insert the Facebook pixel ID in the settings page and it is connected.

Assuming you do not have your pixel setup, to set it up, you go to business.facebook.com, enter your business manager or create on if you haven’t yet, then enter the “Events Manager” page.your ads manager page, click the business tools button, then events manager. This takes you to the pixel creation page where you can create your pixel if you do not have one. For SmartBot360 chatbot purposes, you simply copy the pixel ID that is provided when you create your pixel, or you can look in the settings tab.

Once you do that and save the pixel ID on the SmartBot360 platform, you will then be able to track chatbot engagers. You can then add a Facebook Pixel box to the SmartBot360 flow diagram at the position where you want the bot to track the user. Then, choose an event like “contact” or “schedule” in the Facebook Pixel box on SmartBot360. If everything is setup correctly, the pixel should now fire when users reach the step defined in the chatbot

Creating A Custom Audience

After you confirm the pixel is connected, give it a few days for users to engage with the chatbot and reach the point defined for the pixel to fire and track the engagement. Once you have about 10 or more, you can start with a small budget to retarget these people using Facebook Ads and their platform.

When you see events tracked on the pixel, you go back to the pixel page, and click the green create button on the top right and click create a custom audience. Then choose the event you defined for chatbot engagers, name it, and then confirm to create your chatbot engagers custom audience.

Creating Retargeting Ads

After creating the chatbot engagers custom audience, you then go to ads manager, create a new campaign, and add the custom audience you created in the custom audience section on the adset level, and that is all you need to retarget chatbot engagers. Or if you have trouble creating the custom audience, you can simply do it on this page as well by clicking “Create New” button located above the Custom Audience section on the adset level.

The ads you would create for this audience should be written with the knowledge that they have already visited your website and engaged with the chatbot, so the message needs to be different than one towards completely cold audiences that have never heard of your healthcare practice.

Our chatbot users see a high ROAS on their ads that retarget chatbot engagers because this audience has already engaged more with your website compared to regular website visitors, psychologically causing them to recognize your brand more.

Get more out of your chatbot and website by using Facebook ads to retarget warm and hot audiences.

by Anthony Y. 10/30/2020

With all the advancements in the field of healthcare that the technological revolution of the past two decades has brought, one of the more subtle yet potentially game changing developments is the introduction of chatbots. A chatbot is nothing but a simple program designed to engage audiences in a textual conversation not unlike a human would. Chances are you’ve already come across one without realizing it, for more and more healthcare providers are calling upon the services of patient engagement chatbots to streamline their value delivery funnel!

The typical scenario often looks like this – upon experiencing recurring symptoms, a potential patient logs on to the website of their preferred healthcare provider. There they’re met with a chat window asking for their symptoms and (in some cases where the patient engagement chatbot is HIPAA compliant) their medical history. Once those details are entered, the chatbot indexes the information against its data repository to match it with reasonable conclusions and presents the patient with its analysis. Following the analysis, the user is typically given the option to schedule an appointment with their practitioner along with whatever other necessary steps might be needed. Now, none of these activities are different to what an actual human healthcare provided would have engaged in, but what’s astounding is that they are performed by a few lines of code with the same degree of efficiency and accuracy! The most important quality of a chatbot in the healthcare space is its ability to engage the patient, and as witnessed by this use case example of a HIPAA compliant chatbot such as SmartBot360, you can see exactly how this patient engagement is achieved.

As our little “case study” must have illustrated, there is a world of benefits associated with patient engagement chatbots, but the most important one is how it simplifies the entire process for both healthcare providers as well as patients. There are four primary dimensions to this process of simplification, and the first and most obvious one is how for the patients there is instant engagement. In cases where the symptoms are often very mild, a patient engagement chatbot cuts through the hassle of having to schedule an in-person appointment with the practitioner for what may very well be a non-issue, as well as having to deal with the delays of trying to block time for a phone call at the least. A patient simply needs to fill in the required details with the chatbot and they can get their answers immediately. The efficiency increase that this brings about on the part of the healthcare provider leads us to our second dimension, which is that patient engagement chatbots enable healthcare providers to create more value for their customers. Since the chatbot takes up tasks that are fundamentally repetitive and mechanical in nature, it frees up the human resource in the value creation process to allocate its finite energy towards other parts of the process which not only allows the healthcare providers to serve their patients better, but also improve their bottom line in the long run. The bottom line improvement is indeed the third dimension of how patient engagement chatbots simplify the healthcare provision process since due to their easy to set up and use nature, they can often be used in a plug and play manner thanks to the high degree of specialization and within minutes of being set up these chatbots can start bringing in sales and maximizing patient conversions.

That last part is particularly important since the fourth dimension is how one of the major ways patient engagement chatbots simplify the process of healthcare provision is by providing a platform for setting up appointments with healthcare professionals in a jiffy in a contactless manner. In today’s COVID-19 ridden world where minimizing contact is being advised, having a useful tool that screens for the disease and suggests appropriate action without involving human contact in any form is of paramount importance.

As big a part of the healthcare industry as patient engagement chatbots might be today, they are only going to get bigger and better from here. With advancements in the fields of artificial intelligence, machine learning and linguistic programming, we are getting closer and closer to chatbots that are successful at replicating the human interaction a patient might have with a high degree of accuracy. Given how our world is only going to get more and more digital from here, having our preferred practitioner’s chatbot show us a more human touch doesn’t sound too bad.

by Vagelis H. 10/28/2020

One of the most common questions I get from potential customers is how smart the chatbot is, or if it is able to respond to free text. In this article, I will explain how SmartBot360 tackles the problem of chatbot intelligence, by offering multiple solutions, which can be combined to create a robust chatbot experience.

The solutions that I will describe are available to all SmartBot360 accounts, including free trial accounts. In most cases, a SmartBot360 agent helps set up these solutions, as some of them can get tricky.

1. Multiple Choice (or Yes/No) with no option for free text input.

Here is how to achieve this behavior in SmartBot360’s Bot Builder.

This is the simplest choice, but surprisingly popular for many applications. Not checking the checkbox is a bad idea if the chatbot is not set up to handle free text, as shown below, because the chatbot would just respond “Sorry I didn’t get it”, which constitutes bad user experience.

2. Text-Match: Keyword Containment

The next level of intelligence is to check if the user’s utterance matches or contains a given list of words or phrases. In the below example, if the user’s utterance is “insurance” (which is also a quickreply as shown on the left) or contains the words “anthem” or “aetna”, the flow goes to the top message box.

3. Approximate Text-Match

The next level is to expand to approximate matches. SmartBot360 supports 3 types of approximate matching, as shown below: 

1. approximately equals: checks if the user’s utterance is just a couple of characters different from the condition, for example, if the user types “insurace” instead of “insurance”.

2. approximately contains: checks if the user’s input contains a substring that is approximately equal to the condition. For example, the user says “I want to make an apointment” approximately contains “appointment”.

3. semantically matches: this uses AI (specifically, deep learning) to check if the meaning is similar. For example, the user may type “university” instead of “college”. Note that this works with whole phrases too, not only a single keyword.

4. Use a “catch-all” set of questions to catch unmatched user utterances.

Consider the above diagram. If the user types “What are the hours?” the bot would normally say “Sorry, I didn’t get it”. SmartBot360 allows to link the chatbot to a list of questions and answers (FAQs), which are searched when the diagram does not have a match.

To enable this, one has to first upload a list of FAQs, creating an FAQ chatbot, and then link this FAQ chatbot to the current flow diagram chatbot, as shown below.

Read more details on how catch-all works.

5. Switch to chit-chat if everything else fails

Continuing the above example, if there is no question the FAQ bot related to “What are the hours?” then instead of saying “Sorry I didn’t get it”, SmartBot360 offers another option, which is to have a “chit-chat” engine respond (see above picture). For example, the response may be “Why do you want to know” or “Keep going” or something else. These responses are not useful but their goal is to keep the user engaged to minimize drop-outs.

6. Document Search

This feature is currently being released by SmartBot360. Specifically, a chatbot may search a set of documents (for example, the pages of a web site) for relevant content to respond to a user utterance. This feature is more relevant to customer service applications.