Mon | Nov 19, 2018

Health + Tech | The ‘new oil’: Data mining

Published:Sunday | September 24, 2017 | 12:33 AMDoug Halsall

In this digital tech-centric world, data is king and it is time we, as a country, start mining data to underpin critical decisions. 

At its most basic, this is described as analysing a large amount of data to determine patterns and trends that may exist, with a view to finding solutions, forecasting and predicting future possibilities. 

Data mining can be one of the most profound tools used in any industry that seeks to understand clients and customers’ needs, wants, dislikes and any other pattern that they may establish. 

Companies like Amazon and Google, use data mining, quite successfully, to improve their business processes and offer wider choices to their customers. 

Importantly, they also use this to predict what future customer needs may be and take action by developing and improving their offerings. This is how data mining works – the data is analysed, predictions are made and insights are actioned. All stages are important to maximize the gains from having access to this data. 

Data is a very valuable commodity to any industry and is increasingly referred to as the "new oil". Data mining is particularly important in the healthcare industry which can be revolutionized if policy makers are able to observe trends over time, successfully predict patient and likely disease behavior in a particular setting or community, at a particular period and put policies in place to mitigate likely issues.

A data-driven, decision support approach to solutions in healthcare has to be at the forefront of our attempts to resolve long standing health issues and put policies and structures in place to address them. 

Policy makers can use this kind of information to determine which programme should be funded when, the best areas of focus for a particular period, guide things like drug acquisition and even forecast what professions in the healthcare industry will be more relevant in the future.  This enables more efficient use of our already scarce resources. 

The passing of the National Identification System Bill recently in Parliament is a positive move towards supporting the eco-system that will make data mining in the healthcare sector more granular and therefore more valuable. 

The eco-system that enables linkages across all facets of healthcare including electronic medical records, insurance adjudication, pharmacy, lab and diagnostics already exists.

In fact, the University Hospital of the West Indies (UHWI) is the first in the island which will be fully digitised and therefore in the best position to mine data from these already existing linkages. 

The system at UHWI is so decision support data-centric that it even can discern genetic trends by storing the linkages of three generations.

Professor Archibald McDonald, perhaps because he’s coming from Dean of the Medical School and a doctor himself, placed great importance on predictive data and selected a system that will identify genetic trends.

In addition, for increased efficiency, government hospitals could, through this digital eco-system, ensure that a patient does not go to all facilities for the “same” ailment, thereby receiving multiple prescriptions and wasting resources.

This data can also assist in determining the best locations for diagnostic and treatment centres. At the height of its opioid crisis, the State of Indiana used data mining to determine where to place its treatment facilities to capture those most in need. 

We could use the data in a similar fashion with our limited budget to maximize how we provide treatment and prevention services to our population. 

Data mining can identify health issues in communities before they reach crisis levels; how they move across the island, who are most affected, and catch a potential issue before it can become an outbreak. 

In the cases of the Riverton fire and CHIKV a few years ago, health insurance carriers were able to identify the trends in medication and thereby predict where, and the level at which, persons were affected by both incidents. 

For any health issue, data can determine transmission patterns and recovery rates as well as indicate the population and geographic location that is most vulnerable. 

The National ID system, if incorporated with our already existing technology, can provide data on an individual from the first vaccine or interface with the health system to their last.


- Doug Halsall is the chairman and CEO of Advanced Integrated Systems.

Feedback: or