Big data use in pharmaceuticals has been spreading quickly over the past few years, bringing with it some exciting developments. This data has generally come from healthcare outcomes, clinical trial results, national health trends, and internal data, but is it time that pharmaceutical companies started looking outside of these areas to help create more diverse and robust datasets?
One of the key elements to this is going to be gathering data from a wide range of sources, many of which may be outside the traditional range of data sources, such as social media, environmental data, or even weather data.
It may sound a little off the wall to claim that pharma companies should be looking to Twitter, the EPA, and the sky to get data to help their business model, but with the wealth of data available to them outside of their traditional silos, it would be foolish to ignore them.
One of the key elements to any pharmaceutical development is regarding the enrolment and undertaking of clinical trials, which is always one of the most time and resource intensive elements of any product launch. One of the major difficulties is in actually finding people to undertake them, with Forbes claiming that 30% of the time spent on clinical trials is on recruitment and almost 40% of trials miss their targets. Using data mined through social media may help companies to find people more easily, either through targeted advertising or looking for specific kinds of people or those who suffer from specific diseases. This could save a huge amount of resources in this process and help to complete more thorough clinical trials in a shorter amount of time.
However, although this is undoubtedly powerful in terms of saving money and resources, the ability to use data to predict trends is arguably going to have considerably more impact.
The opportunity to look at general trends within social data is another important aspect, potentially helping to identify gaps in the market or new opportunities for iterations of existing products. This could be through investigating specific areas, sub-groups related to specific product types, or just simply monitoring what people are saying about the products that already exist. This may sound relatively basic and when you take this kind of messaging in isolation it is not necessarily too useful, but once you can analyze several messages over time, trends become considerably clearer.
Pharma companies can also look to more diverse data points that may not even directly relate to customer opinion, such as weather forecasting.
Having the ability to predict the needs of populations needs to focus around understanding the elements that may be impacting them at any one point in time. For instance, if the weather is particularly hot there is likely to be an increase in insect bites, burns, and dehydration, whilst in cold conditions, the chances of colds, flu, and respiratory problems are increased. Having the ability to predict these well in advance, either by looking at general annual trends or monthly changes, can give pharma companies a considerable advantage over competitors.
Demographic data can also offer a considerable advantage to companies by focussing on populations within specific areas. If they know that a certain country has an ageing population then they can concentrate production of products aimed at older generations, whilst a population with increased birth rates can focus on pre-natal or infant-focussed products. If the data can become granular enough it may even be possible to filter down to local levels and focus either marketing or supply to smaller and smaller areas, making campaigns more effective and managing supply levels.
Each of these (plus the thousands of other data sets available) taken in isolation can give a slight advantage to a company, but what will truly make a significant difference, and the element that big data excels at, is bringing all of these disparate data points together to find correlations within the data. It is in this area where the real jumps in performance will come, for instance, with the data above you may see that during a hot dry spell in a specific town, that those with children under 5 are more interested in a specific product or service than at any other time of the year. This is where true value is created, in being able to notice patterns throughout correlated data that can bring genuinely actionable insights. It doesn’t necessarily sit within the traditional data sets for pharma companies, but just like any other company, their job is to create a product to solve a problem, through using a wider range of data they will have the opportunity to find more problems and solve them.