The debate around AI for healthcare decision-making

The COVID-19 pandemic caused enormous pressure in the healthcare sector and exposed its lacunae. Digital transformation in healthcare stepped in as a ray of hope and boosted the efficacy of healthcare professionals who embraced technology. 

Artificial Intelligence (AI) became the favourite technology of healthcare service providers because of its power to revolutionise nearly every facet of healthcare. Hence it comes as no surprise that 94% of healthcare executives say that they have seen a pandemic-induced, increased AI deployment in various departments in their organisations.

Some of the areas where AI made a foray into healthcare are:

Remote patient monitoring:

AI facilitates monitoring the progress of a patient's health remotely, and this helps detect ailments at early stages. This enhances the quality of patient care.

Administrative tasks:

Healthcare service providers are successfully deploying AI to manage routine, time-consuming processes to boost their accuracy and speed.

AI-powered chatbots:

These chatbots guide patients through crucial stages of their healthcare journey, right from scheduling appointments to resolving their queries.

AI-based analytics:

Analytics facilitate a data-driven approach to healthcare. They pave the way for precise diagnosis and customised treatment plans thereby boosting patient experience.

Drug discovery: 

Research indicates that “AI applications in drug discovery are expected to reach $3.5 billion by 2028.” This is because of the attributes of AI like improved accuracy, reduced timeframes, and cost-efficiency, among others. These accelerate the development of new drugs at a fast pace reducing the burden of chronic illnesses.


AI-assisted surgery enhances the precision, accessibility, and efficiency of surgical procedures.

Medical imaging:

AI has revolutionised medical imaging with precise and swift image acquisition, analysis, and interpretation.


AI streamlines radiology with better diagnosis, monitoring, and treatment. Machine Learning (ML) deployment has increased by 60% among radiologists in the past five years.

Clinical decision-making:

AI has made inroads in the realm of clinical decision-making assisting physicians and doctors in making informed decisions. Data reveals that approximately 80% of executives working in the sector will see major changes, over five years, in the clinical decision-making methodology.

Besides these changes, AI is making huge strides in areas like AI-assisted medications and prescriptions, pharmaceutical manufacturing, and more.

Hence, we can see the overwhelming promise that AI holds for the healthcare sector. But can we rely on AI completely for making decisions in different spheres of healthcare?

After all, AI works on a binary representation of problems and solutions

So, is it capable of profound decision-making that is interlinked with multiple factors?

Let’s look at the limitations of AI in clinical decision-making:

  • The success of ML depends on the availability of massive datasets. This could be a problem in the healthcare domain where often patient data is considered confidential and not shared. Therefore, ML algorithms will be less extensive. While they may be effective for individual care, they are limited in scope for improving healthcare.
  • ML algorithms make predictions based on past data and these may not be accurate for specific conditions. This is because these algorithms learn the interrelation between certain patient features and outcomes. Hence, if there are too many variables, the algorithm makes inaccurate predictions. Besides, if any inaccurate data enters the dataset, ML will make inaccurate predictions.
  • ML algorithms have the bias of the programmer and neural networks work on existing stereotypes. Hence, it is likely that the training data may not be inclusive and cover diverse communities resulting in distorted outcomes.
  • Deep learning algorithms lack the ability to justify their forecasts making it difficult for scientists to comprehend the relation between data and forecasts.
  • Most of the AI research in healthcare has been conducted in non-clinical settings. Hence, there is no evidence of its impact on patients. Therefore, generalising results may be challenging.

Besides these, AI-based healthcare is plagued with issues such as data privacy, social issues, ethical issues, hacking issues, etc. AI has great potential to enhance the quality of patient care and patient experience. However, certain measures can be taken to address some of these issues.

Every new AI-based application must begin with explicit questions and discussions with clinicians. These questions must be solved with accurate datasets. The datasets must also include concealed variations for accurate outcomes. Domain knowledge is a must while programming the AI bot. It is also necessary to make the relation between a given input and expected output causal and unambiguous.

While such measures will resolve some of the shortcomings of AI-based systems, AI is not capable of replacing healthcare professionals as of now. This is because healthcare service providers bring with them clinician autonomy, experience, and judgement that cannot be replaced by AI. 

The healthcare domain is not as straightforward as manufacturing or other standardised sectors. This is because clinicians need to work on the uniqueness of each patient and the complexities of individual health conditions. This makes it difficult to standardise the treatment processes. However, AI offers several benefits to the healthcare sector and is a must. The solution is to be aware of its shortcomings and use it judiciously.

*For organizations on the digital transformation journey, agility is key in responding to a rapidly changing technology and business landscape. Now more than ever, it is crucial to deliver and exceed on organizational expectations with a robust digital mindset backed by innovation. Enabling businesses to sense, learn, respond, and evolve like a living organism, will be imperative for business excellence going forward. A comprehensive, yet modular suite of services is doing exactly that. Equipping organizations with intuitive decision-making automatically at scale, actionable insights based on real-time solutions, anytime/anywhere experience, and in-depth data visibility across functions leading to hyper-productivity, Live Enterprise is building connected organizations that are innovating collaboratively for the future.

Recent Posts