[Opinion] How AI can help with customer service in healthcare?

This article comes from Pranay Jain is the CEO and Co-founder of Enterprise Bot. He founded the company alongside his wife, Ravina Mutha, and CTO Sandeep Jayasankar. Enterprise Bot is enabling enterprises across the globe to provide an omnichannel experience to their customers by building AI-powered virtual assistants across chat, email, and voice channels.

Artificial Intelligence (AI) in the field of healthcare is expected to grow exponentially over the next few years at a CAGR of 43.5%, pegging its value at over USD 27 Billion by 2025.

With the help of advanced variations of the AI tech, such as Computer Vision (CV) and Machine Learning (ML), massive strides around the fields of precision medicine and disease discovery have been accomplished! Analytical models are currently being deployed to study vast amounts of clinical notes and data, which will help to identify anomalies that were overlooked due to human error. 

What about drug discovery? On average, around USD 2.6 Billion is poured into the development of a new drug, where a large portion is sunk into clinical trials and overcoming regulatory red tape. With the help of advanced AI algorithms, research facilities are circumnavigating this expense. Leading bio-pharma company Pfizer is now deploying a renowned Machine Learning platform to search for immuno-oncology drugs. 

However, does the impact of AI and ML stop at just pathology? What about other areas across the patient care journey? 

When the global Coronavirus (SARS-CoV 2) pandemic broke out, it became evident that global healthcare systems required more than just advanced medical care. In a recent survey, it was observed that 34% of top healthcare providers hoped that their investment into AI-based solutions would help them make processes more efficient, while 26% hoped to reduce operational costs significantly. 

Due to poor operational procedures and mismanagement, the losses across the healthcare sector have mounted steadily with the onset of the pandemic. Between July to December 2020 alone, the United States healthcare ecosystem lost around USD 120 Billion, taking the total loss to USD 323 Billion for the entire year. 

What went wrong?

To curb the spread of the virus, industries and workplaces had to usher in the remote-working (WFH) model. For occupations that couldn’t work remotely, such as healthcare, it was mandated for employees to social-distance themselves from each other to prevent the spread of the virus. Whilst clinical professionals were required round-the-clock across the front lines, backend operations of medical institutions were made to endure the burden of working short-staffed. This led to an increase in administrative backlogs, coupled with patient care mismanagement. 

However, the problems of global healthcare engines existed before the arrival of the COVID-19 pandemic. The shortage of skilled non-clinical human resources plagued healthcare providers for decades. 

Even though the global medical field was poised to add over 40 million new jobs by 2030, excerpts highlighted that a shortage of over 9 million personnel was inevitable.  

This was attributed to a slew of pre-existent reasons, such as improper hiring practices and lack of workforce availability. All of this directly contributed to poor patient care management. 

In such a scenario, medical institutions had to scavenge for alternative solutions to help enable better experiences across the patient care journey. 

The answer lay in automation. 

In the field of medicine, Electronic Health Records (EHRs) play a pivotal role in diagnosis, pathology, drug delivery and medical history analytics. Since EHRs function as a single source of truth for all patient-related data, it takes up a lot of physician and administrative time. The American Medical Association had disclosed in a study that 70% of doctors and office executives spent over 10 hours a week on just routine paperwork. A majority of which includes manually entering clinical notes into EHRs. However, most of this data goes unused. This changes with the arrival of intelligent virtual assistants. 

Intelligent virtual assistants utilise the Natural Language Processing (NLP) component of their backend to provide an accurate and medically precise speech-to-text functionality. Physicians can now vocally record their clinical notes in real-time during a consultation, and the NLP component converts it to textual verbatim. This text is then automatically stored on the EHR and updated across the HL7 channels. Hence, doctors can now eliminate mundane paperwork from their scope of work, helping them focus more on patient care. 

One of the biggest roadblocks across the patient care management journey is seamless access to medical history. Most patients are clueless about how they can access their data and what the data meant for them. 

Studies showed that most patients found it difficult to communicate with virtual assistants due to a multitude of fears. Either they were scared that the digital assistant was misusing their medical data, wasn’t adept at safely handling the data or found it impossible to communicate with a monotonic AI.

Conversational AI solves this problem. With the power of Meta-Learning, NLP and Conversational AI, intelligent virtual assistants can now converse with patients in more humane and empathetic tonalities. The assistants parse through textual or vocal interactions to create a profile (based on gender, age, nationality, etc.) for more efficient communication. This form of aligned content puts the patient at ease, allowing for a smooth exchange of data. 

However, there is an important factor you need to consider when we talk about handling patient-related data. When the HIPAA Act came into existence in 1996, healthcare providers had to handle patient data with utmost care and security. Lapses in handling this data can expose the establishment to libel charges based on criminal negligence. Hence, any digital assistant must be HIPAA compliant to be deployed across the healthcare ecosystem.

With the arrival of COVID-19, personalized care delivery became skewed as healthcare professionals were working around a virus that they had not seen before! 

For example, a person with a history of pulmonary issues is prone to breathlessness and sudden drops in Oxygen-Saturation levels. Following COVID-19 protocols, any person with breathlessness and Oxygen-Saturation levels below 94 was to be administered COVID-19 treatment protocols. Such situations became common, and at times have turned critical. 

This could have been avoided if doctors and clinicians had access to medical history in real-time. 

Enter healthcare virtual assistants. These assistants are integrated with front-desk operations to ensure that a patient’s medical history is retrieved from the EHR in real-time to help physicians arrive at an accurate prognosis. 

Is customer service limited to customer interactions alone? Not in the healthcare industry. An important facet, which primarily occurs in the backend, is the quality of care imparted to patients. A side of the dice that no one sees. 

Preventive medicine has helped extend the average life expectancy of the human race by helping to detect life-threatening diseases before they turn deadly! In 2017, a group of medical researchers from the prestigious Yale University showcased their studies and findings regarding the role of a family of proteins called Fibroblast Growth Factors (FGF) in blood vessel development. 

Their study displayed how these proteins contributed to tumour growth and other cardiovascular diseases. Using proprietary AI algorithms, the researchers were able to formulate connections between Molecular Functions, Gene Location, RNA-sequence Variations and Expression Levels, which led to the discovery that FGFs would become dominant by controlling glucose metabolism levels.  

Hence, the type of treatment and medication administered becomes more focused and accurate. 

Hospitals rely heavily on building brand loyalty within their operational zones to ensure constant footfall. To create a certain level of trust and dependence amongst individuals, medical establishments need to go that extra step in providing next-gen healthcare.

Clinical aid doesn’t always usually end with singular consultations. And the entire process of managing bookings, follow-ups and factoring no-shows is a dreaded task across all healthcare front desks. Preventing gaps from arising in bookings and booking management ensures maximum profitability as physician time and resources aren’t wasted.

Take the case of COVID-19 vaccination drives in India! Due to the dense population and limited doses of vaccines, central health protocols enforced a rule where hospitals and vaccine centres were not allowed to open a vial without at least ten people present. This would prevent the wastage of vaccines. However, in rural areas (Tier 2 & Tier 3 cities) across the subcontinent where vaccination turnouts were abysmal, people were left waiting for hours before ten individuals showed up.

This very predicament could have been avoided if the hospital had a dedicated virtual assistant that analysed the due dates for second doses, charted it across the calendar based on availability and informed specific individuals on the time and date. 

Hence, it becomes evident that AI in customer service across healthcare systems isn’t just limited to interactions with patients and giving them access to information in real-time. It features in the backend, where it helps scientists and researchers discover more advanced treatment pathways and aid drug discovery. And in the front-end of operations, it’s all about making sure doctors and clinicians can focus more on the patient instead of drowning in paperwork and routine administrative tasks. 

By automating almost every step along the patient care journey, doctors can now focus on patients and ensure they receive the best possible care. You’d be interested to know that this boosts physician productivity, which in turn increases morale and reduces burnout, attrition rates and overall employment displeasure. 

At the end of the day, this is what a doctor was educated and trained for, not paperwork! For that, we have the technology. 

So, let’s leave the grunt work to the AI and focus more on critical thinking.