Big data. Everyone is talking about it, nobody really understands it and even fewer people are actually getting any value from it…most especially those in healthcare. The UK has just had a public debate regarding patient data which offers insights to those with an ear to hear.
It has been estimated that until 2003, 180 Billion people had lived on this earth. Together they generated a cumulative 5 exabytes of data. We now generate that in just half a day! Of course, not everything we Tweet or put up on You Tube is of any public interest. However, when data in multiple databases and legacy systems is joined to information on social media, we have the capacity to generate new sources of meaning and relationships between parties with complementary interests.
The UK is home to a number of pharmaceutical companies which earn export revenue and generate national pride. The government sold Big Data as an opportunity for patient health records to validate drug regimes and treatment protocols. Few in healthcare would disagree with the need for more long-term, evidence based data, especially on drugs in multi-morbid patients. However, data to drug companies quickly degenerated into records equaling record profits. The NHS backed down.
In Australia, we already have a growing national health record. The ability to move data across sectors has in theory been achieved in ways which make our system the envy of the world. Google also has an impressive record of being able to link its search engine metrics with outbreaks of pandemics far earlier than traditional public health systems are able to. If we consider healthcare needs to become integrated with social care to resolve or support long-term conditions, the ability to join up information on patient preferences, mental health and physical health scattered across a range of data repositories offers the enticing possibility of being able to deliver far more timely and customised services than we ever thought possible.
What about accuracy? The possibility of big data is that by gaining mutually confirming insights across a wide variety of datasets you can overcome the inaccuracies of any particular dataset. Volume equals value in this case. This is a very different way of arriving at precision from the usual scientific research protocols the last 200 years of medicine has spent refining.
Of course, gaining access to data also means overcoming the ownership concerns both patients and practices have regarding “their” data. Many practices view patient records as a real competitive advantage. It slows patient movement to new practices. But what are practices doing with the data to identify trends or monitor performance? In most cases very little.
When it comes to diagnosis, it is fundamental that any clinician has on hand accurate individual patient data. When it comes to prognosis, population level interventions to improve the quality of care or predict outbreaks of infectious disease, accuracy at the 90% level is probably more than acceptable. This is what Big Data can easily achieve. It is also what our current approaches do most poorly.
The South Africans have a lovely saying…”when you need to go fast go alone, when you need to go far, go together”. This wisdom really sums up the current policy debate for me. Together our data will provide the solutions we need in the absence of radical increases in research funding, or health and social budgets which are not likely to eventuate given increasing demands on public funds. Let’s consider the benefits and sell it a little more cleverly than they did in the UK.
Tracey Johnson, General Manager, Inala Primary Care