We generate 5 exabytes of data a day. That's approximately 5 billion gigabytes. In other words, that equals to all the words ever spoken by human beings. Our ability to generate and store data is growing exponentially as consumer input continues to increase. With all this data at our finger prints, it's not crazy to think that soon we can be able to solve some of the world's most complex quagmires by relying on this infinite pool of information.
The answer is twofold. Before we can even use this information, we must make sense of it. This is where AI and machine learning enter the stage. These algorithms have been in motion for quite a long time but we couldn't utilize them due to the shortage of data that preceded todays boom. Soft AI or machine learning requires large amounts of data for it to reach its potential. To our advantage, we are currently living in an age of 'big data' where the end user (consumer) is the prime generator, unlike the early 2000s when businesses were both the consumer and generator.
We currently have two of the most prime ingredients required for the utilization of data to solve a variety of problems that span across different sectors, including healthcare, which has a myriad of dilemmas waiting to be solved. This begs the question, how can this conducive environment of big data help solve the opioid epidemic?
The answer lies in machine learning. Physicians currently have to sift through a pool of data about a patient, go through numerous clicks to get into their health records, and finally make 'informed' decisions in order to treat a patient effectively. This is a tough job for a single physician who only has one brain to rely on (they aren't as smart as we think). Soft AI and smart algorithms can do that job more effectively than a single physician while also stepping in as a mediator. For example, these tools can be used to screen for a patient's well being in realtime and relaying that information to a physician so that they can make more informed decisions without having to spend too much time, which decreases productivity and response ratescfor both parties. AI isn't meant to a replace a physician's job, rather, its value comes from its ability to make sense out of data points/trends and relay meaningful information to a stakeholder so they can continue doing their job, which is to look after a patient's wellbeing. Furthermore, no machine, as of today at least, can replace the empathy that a human possesses. This fact is equally important because healthcare is an industry that requires a greater attention to humanistic qualities, such as empathy and compassion, which unfortunately has been neglected in the past.
Currently, we are at the intersection between humanity and technology. This environment is conducive to solutions because of the multifaceted approach that it embodies. Solutions that are based on big data and patient centric care can give physicians and stakeholders a conduit to early intervention and support, two factors that have proven track records of success. If we had the ability to know if a patient was going to misuse their prescription again, would we not intervene? This data-centric foresight is crucial in a sector where the knowledge gap between a patient and a care provider is undoubtedly wide. I have heard countless stories of patients who were left in the dark because of a lack of shared information, and not necessarily empathy or support.
Throughout the next few months, we are going to witness a massive paradigm shift in what's been said to be one of the most antediluvian sectors. Industry leaders, institutions, and government officials are beginning to consolidate their efforts to address the epidemic using big data and patient centric care. It's fair to say that they've made great strides in aggregating a pool of data that was once disparate. However, the next step would be to focus on generating data at the point of output. What we currently have is a macro outlook on the epidemic, which includes PDMPs and overdoses. However, they lack complete narratives. There is no beginning, middle, and end, just an end. We need to understand it from a patient to patient perspective and pool together trends that span across demographics and geographies. By focusing on data that is consumer based, meaning directly coming out of a consumer's behavior and not a state/nation wide level, we can begin making sense of when, why, and how misuse progresses into addiction, bringing us a step closer to completely preventing it in the first place using predictive analysis-AI.