Healthcare industry in US is constantly changing. No doubt about that. It is a paradigm shift from a typical “I provide service and and get paid” to “I improve population health and then get paid”. The idea behind this change is simple. Payers have been rolling out huge chunk of money only to see the quality of healthcare not improving in the same proportion as the cost. So to improve the quality of care, the elegant change was brought to life. This is what I think.
This elegant change has led to open questions from all around the industry as a whole. How does one track the health of a population? As a healthcare provider, what if I am situated in a disaster hit area and the population I am treating is not as same in healthy terms as other part of the country – I would end up losing money in the new model. Questions like these have been popping up and are still around the healthcare domain. To answer these questions, CMS is doing a wonderful job in taking questions from the industry and answering them. This just goes to show how healthcare industry right now is in transition.
Accountable care organizations(ACO) have taken over the mantle to reduce operating cost and provide a value based care to improve the quality of care. For them, it is important to figure out the major cost heads, the right set of practices, payer contracts adherence among many different questions. They are in charge of keeping an eye on the measures and guidelines set forth by the government. If the measures fall in the lines specified by the government, the ACO’s attract incentives. All in all, improved quality of care and improved utilization equals dollars.
Now coming to the topic, the headline of the article, data science can answer a lot of these questions. Every player in the healthcare industry needs to build on innovative statistical insight generating engine that will essentially nudge them in keeping a tab on the proceedings and also in making smarter decisions. A doctor, for example should be able to view the entire history of a patient for providing better care. He should be able to view the risk associated with the patients beforehand. An ACO should be able to track the government defined measures to keep a view on the measures for which incentives are linked. They should be able to identify the providers who are performing at par and bring in the best practices in play for the other providers and help them catch up. Innovative solutions to predict the length of stay of an incoming patients will help them manage operations effectively. Knowing the likelihood of a readmission of a patient at the time of decision for discharge would help them keep readmission levels down. So much to do all at once!
Now let us take a view on what is happening in big data space. The word has been around for not so long but has created a lot of noise. To better answer what is happening in big data space let us head back to what was happening in data space a couple of decades back. As far as I can recall, 1990s was when the companies were aggressively trying to shift the manner in which they stored the data. From a paper based collection to a more digital format. The companies hired engineers to develop custom applications to structurally store the data and build reporting engines on top of them. Then came a time in late 1990s and early 2000 when IT companies like oracle, IBM, SAP started to understand the common structure of data and the reporting needed in different verticals. I call it the productization phase of IT applications. This was also the time when we experienced the exponential growth in increasing processing power and decreasing data storage space that gave a massive rise to bulk of data ready to be processed.
The cycle began again! Companies again started to hire engineers to build analytical models in silos and started to gain valuable insights from the predictive models. Now is the time to productize analytics. The analytics industry is on a brink of change.
When we look at it from a broader view, the change in both the industries will compliment each other. Business folks, decision makers, doctors, insurer does not and should not want to deal with team sitting in silos to get an answer. Rather they should have a product in front of them that helps them answer questions on its own. They should have a product that should give them intuitive views to track and monitor their performance indicators. They should have the capability to easily run analyses on a sample of the population.
To summarise, healthcare industry needs a medicine that would help in improving the quality of care. No doubt most of it lies in the hands of physicians, after all they are the ones who actually prescribe medicines. But to look at a population as a whole, one needs big data science to come up with an overall view which can be deep dived to an individual view in a matter of seconds. A wholesome medicine that can cure the age old problem of extracting right information at the snap of the fingers. A product to ingest wide variety of datatypes, bring in data from public domains, act on it, analyse it and then prescribe and visualise it for individual stakeholder, at the same time giving them the power to play around, build and test their intuitions with ease.