Artificial Intelligence is known to have the ability to perform functions like human beings. There are various fields in which AI has revolutionized the way things are done but healthcare is an arena that has been highly impacted. The Indian government has taken a lot of initiatives in boosting AI in healthcare so that rural India can get quality health care. AI and machine learning are making screening of diseases easy and detection at an early stage. Similarly, in pharmacy, it is helping in optimizing the supply chain. However, challenges like lack of regulatory body, effective laws and so on are still looming over the government of India. Ayushman Bharat, National Health Protection was launched by the Indian government to provide quality and affordable healthcare to all its citizens. It is world’s largest healthcare scheme but challenges like database maintenance, quality and so on is on rise. In order to deal with such threats government of India is deploying AI and machine learning and various research suggest how further it can be refined by advancing more into AI.
Artificial intelligence (AI), defined as intelligence that is demonstrated by machines, in other words, mimic human cognitive function (Jiang, 2017). In recent times, AI has been used in many fields like defense, healthcare, banking and so on. In order to provide healthcare and fight against factors like shortage of professionals, non-uniformity access to healthcare, concentration of healthcare facilities in Tier 1 and Tier 2 cities and alike (PwC, 2018). Indian healthcare by 2020 will worth US$280 billion thus, is one of India’s largest sector in terms of employment and revenue (IBEF, 2019). Some of the growth drivers are increasing income level, rising diseases due to lifestyle, increasing level of health awareness and easy access to insurance. Under Ayushman Bharat, one of the largest healthcare program funded by the government of budget US$ 887.04 million for the year 2019-20. The program targets to achieve 500 million beneficiaries and as of February 2019, approximately 12 lakh people have received treatment under this Yojana (india.gov.in, 2018).
Artificial Intelligence in Healthcare
AI has disrupted a lot of industries and healthcare is one of the beneficiaries of the revolution. The areas where AI has been emerging include machines that can sense, comprehend, learn and act in order to execute administrative and clinical healthcare functions (Radick, 2017). AI increases the scope of activities that can be done by machines like natural language processing, chatbots, computer visions, machine learning and so on. Machine learning can be deployed to understand the overwhelming healthcare data thus, reducing evidence-based decision (Westgate, 2017). IBM’s Watson is used in Oncology treatment to prescribe the treatment which is best suited for patients (Reddy, Aggarwal and Acampora, 2015). New startups in India and integrated communication technology treatments are using AI to address challenges like delivering high-quality healthcare facilities in India, automatic diagnosis, detection and screening of diseases, predictive healthcare diagnosis and so on.
Some of the factors that are bound to be affected in a positive manner by integrating AI in healthcare are cost, quality, efficiency, and reach of healthcare to the underserved areas. The major focus is on serving those areas where infrastructure is not available or primary healthcare is questionable. However, technology replaces doctors is still debatable due to clash of interests of different stakeholders. Thus, AI in India is focused more on addressing issues related to economic disparity. Some of the initiatives taken by government include tie-up of NITI Aayog with Google to encourage AI research thus contribute to a bigger cause that is technology-empowered NEW INDIA (ICT Monitor Worldwide, 2018). Furthermore, NITI Aayog is working at national level in deploying analytical portal with the help of AI. India aims to make National Health Stack (NHS), a centralized health record which will help in managing health information effectively by use of big data analytics and machine learnings (egov, 2017).
There are lot of opportunities for AI in India but challenges need to be dealt. Proper advancement is possible only when there is a robust legal framework. Some of the challenges are regularity authority, appropriate certification mechanism, infrastructure, investment, unanswered legal questions, and information asymmetries and perceptions. Currently, India lacks a regulating body concerning AI in healthcare. It is important to safeguard all the data and ensure that data are not misused (Stead, 2018). Data produced by AI needs certification, confirming meeting certain basic parameters. However, India devoid of any such mechanism, lead to expensive clinical trials which is time-consuming and generation of any standard of certifications which leads to unacceptability of data (Stead, 2018). Therefore, possible solution could be to have doctors on panel while deploying AI.
Infrastructure which is of utmost importance to execute AI like cloud computing, high-speed internet and so on are still unavailable in many parts of India (Srivastava, 2018). Hence, most of the servers are outside India. AI is under-funded in India especially by government and currently limited to only few fields. Information asymmetries and perceptions are another areas of concern. Doctors and coders do not have any specific standards to meet the uniformity as a result doctors are hesitated to adopt software. Moreover, understanding of technology still differs as compared to developed versus developing countries. There are lot of unanswered legal questions like who will be responsible in case of any error doctors, coders and someone else (Gupta and Kumari, 2017).
Therefore, it is utmost important that there should be strict laws and orders in place to answer all the challenges before AI is rolled out in any fields especially medical.
Application of AI in Healthcare Segments in India
Hospitals in India have witnessed employing descriptive and predictive AI. Manipal Group of Hospitals in order to help doctors in providing better treatment to cancer patients have IBM Watson for Oncology software installed. With the help of AI, it analyzes data and provides research-based evidence thus, improving quality of decision making. Another example of AI being used in detecting diseases in Aravind Eye Care where data collected is analyzed for early detection of diseases like diabetes, blood pressure and so on (Vedakkepat, 2015). In pharmaceuticals, AI can be helpful in enhancing the value proposition, automating sales, differentiating it from competitors and so on. Pharmarack is utilizing AI in automating the supply chain management (eHealth, 2019).
In the area of diagnosis, both startups and well-established companies are harnessing the potential of AI in diagnosing diseases. Some of the big names in the field of AI healthcare include Google and IBM. However, lot of startups are using deep learning, machine learning to detect the symptoms of diseases or recommend personalized care. Cureskin diagnoses various skin diseases conditions like scars, pimples and so on. Qure.Ai uses deep learning which enables in detecting as well as recommending personalized treatment. Another area where AI is extremely helpful is seeking help while facing depression. Depression is seen as a stigma in Indian society. By use of AI, it helps in providing chatbots (Wysa) which give empathetic support and suggest when to consult practitioners. Since, identity of the person is masked people usually open-up and get the advice.
Telemedicine helps in providing quality and affordable treatment in rural India. Since, human part is eliminated AI helps in standardizing the quality of the treatment (Chandwani and Dwivedi, 2015). SigTuple analyzes blood sample and generates reports without help of pathologists. Others include Philips Innovation Campus (PIC) which is enabling healthcare to become affordable and accessible by use of AI. Philips by partnering with Fortis Escorts Heart Institute, Delhi has come up with IntelliSpace Consultative Critical Care to monitor different intensive care units form a central command center.
Ayushman Bharat (NHPS)
India’s population is about 1.3 billion which is just a shade less than China’s population. India one of the main challenges is to provide quality healthcare to all its citizens. Though India’s major population that is 65% is young and under 35 years old but has major chunk which is aging and needs quality, affordable healthcare system. Indian government launched one of the world’s largest health schemes covering approximately 10 crores poor and 50 crore vulnerable families providing coverage up to 50 lakhs rupees per family for secondary and tertiary care, Ayushman Bharat- National Health Protection (Ali 2018). It was incepted in 2018 under Ministry of Health and Family Welfare.
According to Lahariya (2018), some of the key features of Ayushman Bharat are described below.
- The scheme will have specified coverage of 5 lakh per family per year.
- It is transferable in nature across the country. The beneficiary can avail cashless coverage benefits from any empaneled hospitals across the country.
- The National Health Protection Mission is based on entitlement which is decided based on the SECC database.
- Both public and private hospitals are listed from where beneficiaries can avail the scheme.
- In order to restrict payments for treatments, a pre-defined package is made.
- Core governing principle of Ayushman Bharat co-operative federalism and state flexibility.
- To have robust and scalable IT platform, National Health Protection Mission has a partnership with NITI Aayog to have paperless, cashless transactions.
The scheme includes Rs 5 lakh insurance which can be utilized not only by individual but by family members. In case of multiple surgeries, the highest package is given followed by 50% waiver for the second time and 25% waiver for the third time. The cashless treatment is shared between central government and state government by 60:40 ratio (Jha, 2018). “The beneficiaries once verified as genuine receives card which can be shown to any impaneled hospitals to utilize the service. Most of the critical diseases are included in the scheme barring few like cosmetic surgery, fertility-related procedures and so on. The aim of the scheme is to provide quality healthcare to all the citizens” (Keshri and Gupta, 2019). However, there are many challenges which Ayushman Bharat faces and in subsequent sections, it is detailed how AI can be used to overcome those challenges.
Ayushman Bharat Challenges and AI Utilization in Solving those Challenges
One of the success stories of AI for Ayushman Bharat was enabling cashless transactions at multiple levels (Bakshi, Sharma and Kumar, 2018). In case of emergency, service patients can go to any hospitals and get treatment. With the help of AI, it was possible to have treatment in any hospital and get information about the patients via online channel. This reduces the stress level of patients, eases the transaction system and helps in getting the required treatment. Also, at the empaneled hospitals because of AI all the transactions are cashless. Ayushman Bharat currently facing challenges like data handling of all the patients and citizens enrolled in the scheme, minimizing the cost of service, data security, quality of care and so on. In order to enhance the service, Indian Government has deployed AI, Machine learning and deep learning. Below are few challenges along with solution on how AI can resolve the problem.
- Minimizing the cost of services at the factor of care.
In India, every year more than 60 million people who are below poverty line are unfortunate to avail healthcare due to its high cost (Angell, et. al. 2019). Currently, it is observed that the Indian healthcare delivery system lacks standardization, as a result, the cost of providing the service is exorbitant. Therefore, one of the most important challenges Ayushman Bharat facing is reducing costs in order to meet the volume and quality. One of the way hospital can reduce costs is using AI in emergency cases where lot of tests need to be done in short period of time (Angell, et. al. 2019). AI not only reduces the time taken for getting the results of tests but decreases the doctors’ time in performing those functions as a result doctors can serve more patients at lower cost and standardization can be achieved.
Online platform and gateway are required to keep medicines, drugs, instruments information available so that it can be accessed easily and by everyone. The supply chain management is important for managing any healthcare organization (Bakshi, Sharma and Kumar, 2018). One of the applications of AI is to automatically send the inventory orders when level of medicines reached certain thresholds while taking into account doctors’ feedback about how the medicine worked on the patients. In case any medicine has negative feedback, it should filter it out and raise alarms so that correct intervention by doctors can be taken and medicine can be replaced, or further investigation can take place.
There are many startups which are or will be legally stationed in medical fields and helping doctors and patient in many ways.
- Ada: It provides a platform which directly communicates with the patients to give recommendations. The application communicates with the patient about symptoms and issues which s/he is facing and in response, it provides proper feedback and offers specialist details for remote consultation.
- Lunit: It is utilizing deep learning and 3D visualization so, that it can be used to detect the unnoticed disease like airway cancer. It aims to perform with 83% to 86% accuracy.
- Sense.ly: This software is similar to Ada but has some different and advance services like reminder for medications and visit to the doctor.
- Insilico Medicine: It was developed in 2014, which performs the most important task in the medical field that is drug discovery at minimal cost and time. The company aims to make a better living of all human beings.
- PathAI (Boston): A graduate of Harvard medical school, Andy Back founded the software to help doctors in finding accurate cell images, which helps to identify the disease to the patients.
- Aira (San Diego): It provides vision to visually impaired people. With the help of artificial intelligence, smart glass helps the person to see the world. For blind people, it can perform some simple tasks like reading so on.
- Julia: It helps in diagnosing diabetic retinopathy, through the proper use of deep learning. It provides proper reports of the respective test.
- Ensuring Quality of Care
With the assurance of quality of care in Ayushman Bharat Yojana, the hospitalization rates of bottom 40% of the population is expected to increase from 2.45% to 2.75% which is about 1.7 crore patients per year (National health authority, 2018). In order to meet quality care of the increasing number of patients, AI plays an important role in improving care. With the help of Artificial Intelligence, quality and cost both get affected, as efficient work minimizes the time period, which reduces the rush at various places like reception, diagnosis centers, medical stores and alike. Though machines play major role in improving the quality of service however accumulation of bacteria and infections on those machines can lead to hospital-acquired diseases (National health authority, 2018). But with the help of bacteria or infection tracker device, which is fully automated, the infection rate can be decreased. The proper alarming system in which the device provides an indication at the point when infection rate is increasing to a certain point can be helpful in monitoring the diseases.
- Quality and security of data
Ayushman Bharat Yojana majorly relies on the development of data which needs to be maintained effectively. The data privacy is another issue that needs to be addressed so it provides security to stakeholders like patience, service providers and so on (Quality council of India, 2019). Proper structure of the data follows some steps like collection, storage, completeness and exchange which ensures the data is reliable and is in accepted format. However, due to lack of proper systems in place, some of the challenges which are faced by the patients include the inaccessibility of their medical records.
Consequently, they fail to present in front of another doctor and case is severe during emergencies. The second challenge is securitization of the data, that is ensuring security of data and its privacy during storage, standardization and exchange among different entities (Quality council of India, 2019). Nowadays, collection of amounts, receipt, storage facility and transfer of sensitive information is subjected to the Information Technology Rules, 2011which is covered under Information Technology Act, 2000 which is shared by the service providers to the respective authorities. Information and data which is shared among the hospitals, diagnostics centers, and clinics need to be maintained properly.
Both challenges can be resolved through deep learning. In the hospitals by use of different software data and information is made secured and accessible. IT heads lead and circulate information between hospitals and authority with the help of programming language R and Python.
- Reducing frauds
Globally, it is noticed that every year $260 billion or 6% of global health care spending is lost in frauds. Health care schemes, insurance policy, and programs which are organized by the Government or through any foundation, never have symmetrical information from the subordinates which leads to the fraudulent activity in the whole system. This leads to malpractices, providers’ persuasive demand, fake policyholders, fake beneficiaries and so on. In order to address these malpractices in Ayushman Bharat Yojana, there is a framework as subjected to Anti-fraud Guidelines, 2018 which aims to detect, prevent to deter fraudulent and abusive activities in the system (Dey, 2019).
The Government of India put an extra layer of protection by using AI to monitor the trend and forming standard treatment protocols to check irregularities due to over-billing or over charging, over-testing, wrong beneficiary information and abuse in referral mechanism. The scheme touched almost 30 lakh beneficiaries within 10 months, the Government implemented an agency, National Health Authority (NHA) for prevention of fraud in the scheme, which continuously detects the potential frauds and does detailed investigation before taking any measures against them (National health authority, 2018).
NHA had detected 48 hospitals across the country out of which 31 got suspended because of the following reasons:
- Irregularities of the doctors in public hospitals during their working hours.
- Public hospitals doctors were illegally referring to the patients in the private hospitals by issuing the referral slips for pecuniary benefits.
- Referring from doctor’s own clinic to the private hospitals.
- As per the scheme patient who admitted in the hospital for 6 hours and more is covered but, in some hospitals, information was falsified.
Because of all these irregularities, NHA designed an IT system with the help of machine learning which is applied to every hospital and diagnostics center to access real time information. It includes activities like beneficiary identification, fund flows, transaction management, claims payment, referred patients details etc. Through this NHA maintains strictness in system for tightening the noose around the private, public hospitals and insurance companies in which they ensure the smooth function of healthcare benefits to the beneficiaries.
Artificial Intelligence, machine learning, and deep learning bring lot of opportunities in all the fields, but healthcare is one of the most prominent areas where AI can be helpful in bringing standardization of the process, cost reduction and so on. Government intervention has further enhanced the AI ecosystem in providing opportunities for new startups to address the new venture while allowing well-established technology firms to collaborate with hospitals to deploy AI. However, challenged like lack of proper regulatory body, certification of the software have made doctors doubt on the result. Hence, in order to increase the level of acceptance government need to address the above issues.
AI is being used in hospitals to detect diseases, pharmacy supply chain management and many more. Ayushman Bharat, one of the Indian government flagship schemes to provide healthcare insurance to all the citizens of India is an area where AI has and will pay a major role in its success. In order to provide quality healthcare to rural India, telemedicine has been deployed, data securitization is confirmed by using AI. Furthermore, government has added extra layer in reducing frauds and ensuring high quality service. In a nutshell, AI will help the healthcare sector in standardizing the process, reducing cost, eliminating frauds and making quality healthcare service available for all.
This article is co-authored by Prof Raul Villamarin Rodriguez, Sanjivni Sinha & Sakshi Tripathi, Universal Business School.