Imagine the scenario where your after-dinner nap gets interrupted by your favourite iPhone ringtone just to hear the pre-recorded voice of Scarlett Johansson (did I mention the movie ‘Her’ is one of your favourites?) reminding you gently that your regular pill is overdue for more than 60 minutes. Scary thought? As a matter of fact technology enabling medication intake monitoring already exists and can be purchased from multiple vendors including AdhereTech and Proteus Digital Health. It may involve physicians to precisely check your adherence, even in real time if necessary.
Unsurprisingly, in our world ruled by overall cost efficiency, it is more likely that Scarlett will be replaced with a pre-recorded message which is carefully personalised for you by a specialist adherence and wellbeing solution provider. As a result, soon after the call you walk to the medicine cabinet (with a touch of unhappiness associated with leaving the comfort of your favourite resting spot) and swallow the pill. Done.
If you find the previous scenario too much sci-fi driven (or generally misaligned with healthcare budgets) there is a plan-b option. How about a personalised medication reminder call (yes, you are still allowed to hear the same favourite actor’s voice, or even better, video transmission in the corner of your latest curved 101” TV) before getting snuggled up? Of course, being privacy-savvy, you would never agree to swallow the pill with a sensor so it may be puzzling how those adherence guardians knew you would forget to take it. Actually, they don’t. They predicted you may not take it under current circumstances so an appropriate intervention was generated on the fly. And it is all driven by data. Even better, a vast amount of intelligently collected and analysed data. Also known as Big Data.
Data management and analytical capabilities follow their own evolutionary path. We often see companies operating in our space typically starting from operational reporting providing endless counts and averages quantifying “what happened”. This may be great for end of month billing and C-level presentations showing results. The step forward may be about “what is happening now” and describes the near to real time forgot-to-take-it scenario from the beginning of this article. Although the sensor technology used for monitoring may be advanced and costly (yes, you will be just fine even if the sensor is swallowed), the actual business impact is limited to a particular kind of event. A little bit trickier is to identify “why did it happen” and many business intelligence vendors thrive in this space. Tools in this area provide rich capabilities but, at the end of the story, it all comes back to analysts and storytellers. Atlantis Healthcare offer specialised data analytics intrinsically as part of a high level service reliability and sophistication required under current pharma marketing conditions. Today’s businesses tend to underestimate the challenge associated with identifying and deploying the right skills in the data science arena (only the US will be short for two million workers with skills in data analysis and management by 2018). Currently our teams of analysts and data scientists combined with health psychologists are capable of understanding “why did it happen” based on a rich combination of available demographics, clinical and digital data collected in Atlantis Healthcare regional data warehouses. All programme learnings are leveraged in our personalised communication strategy with key focus on achieving statistically improved health outcomes. But wait, the story is not finished here.
Clearly disproving the Occam’s razor principle, the area of the highest business impact often comes with the highest technical complexity. In Atlantis Healthcare we are constantly engaged in the process of increasing the capacity of our data warehouses to make them capable of capturing, storing and analysing growing data sets. Patients’ digital footprint now includes streamed data resulting from social network engagements, mobile applications, fitness gadgets, medical devices… The data volume is reaching the level that automation must step in and prioritise valuable data before it hits the data storage farms. Powerful algorithms also need to be trained to remove the noise (in our jargon just the synonym for unusable data) and assist analysts to focus on important facts to discover patients’ behavioural patterns. Once the patterns are identified the statistical learning theory makes it possible to answer the most valuable question for all healthcare stakeholders: “What will happen?”
With behavioural learnings generalised enough for robust models to be built and verified, it is a straightforward engineering step to initiate an appropriate level of customised communication across integrated channels and devices. Even if it requires Scarlett reminding you to take your pill while gently smiling from the upper right hand corner of your ultra-high-definition TV set.
Welcome to the Internet of Things age.
A shortage of two million data management and analytics workers is predicted in the US for 2018 – source: Center for US Health System Reform Business Technology Office, January 2013
World Internet of Things installed base will increase from 9.9 billion in 2013 to 30 billion devices in 2020 – source: Worldwide Internet of Things 2014-2020 Forecast Update (November 2014)
To control spiralling costs, 70% of healthcare organisations will invest in consumer-facing mobile applications, wearables, remote health monitoring, and virtual care by 2018 – source IDC FutureScape: Worldwide Healthcare 2015 Predictions
By 2018, 80% of healthcare data will pass through the cloud as providers leverage cloud-based technologies and infrastructure for data collection, aggregation, analytics, and decision making – source IDC FutureScape: Worldwide Healthcare 2015 Predictions.
