What Big Data Really Means


As big data gains popularity, data evangelists forecast that healthcare enterprises which leverage power of data will develop significant competitive advantage

In this digital era, where knowledge is being increasingly shared across various networks, big data is the new Holy Grail of healthcare executives around the globe. In fact, studies published by consulting firms like McKinsey as early as 2011, states that it can give policy makers and physicians the power to prevent spread of infectious diseases and make treatment more personalised and cost effective.

Big Data refers to studying large data sets in different ways to get valuable clinical and management insights. Industry experts say that it can enable healthcare systems, with complex and large amounts of data, to analyze, understand and predict its own ecosystem. In a data-driven healthcare system, data scientists examine different varieties of data like clinical data, publications, clinical references, genomic data, streamed data, web and social networking data as well as business and organisational data and clinical decisions and business decisions are often made on the basis of these predictions.

So, what does advanced data analytics promise healthcare enterprises? If one delves in to white papers and conferences conducted by various IT companies, the top healthcare applications include personalised medicine, individual clinical treatment based on per patient prediction and analytically enhanced diagnosis. And the best part? It can track population’s health as well as identify pockets and regions that need improvement. Even though it has not trickled in to Indian healthcare industry, experts suggest that it can contain epidemic breakouts like dengue as information can be tracked real time.

Analysts suggest that analysing data can also be major money spinner for healthcare organizations. McKinsey study has predicted that if US health care organizations were to use Big Data Analytics creatively and effectively to drive efficiency and quality, the sector could create more than $300 billion in value annually.
The Golden Goose
In fact, several industries including finance and retail have been successfully utilizing advanced data analytics for some time. “Scientists building speech recognition and language translation software found the potential of data analytics when large corpuses based on the world-wide-web became available- their applications got a lot better quickly, and they were able to throw their complicated mathematical models away,” says San Francisco based Ian Blumenfeld, CEO and Co-Founder of InSample, a SaaS solution for health data analytics. He reasons that we are just hitting the point in healthcare where healthcare systems have enough data sets to start understanding what works and what doesn't in the real world.  This will allow for significant improvement in the treatment of patients. "Imagine if you went to your doctor with a problem and they were able to match you to a thousand others with similar symptoms and phenotype, who were treated effectively. You'd be on the right treatment immediately and to great effect,” explains the data scientist

Data science professionals also forecasts that early adopters, who leverage the power of data will develop significant competitive advantage. “It won’t be long before the only choice will be to take the plunge or go out of business,” warns Blumenfield.

Today, there are eminent IT companies like IBM, Intel and Cognizant which offer technological solutions for healthcare enterprises. An oft quoted example of big data thinking is IBM's Watson. The IT major is working with a number of large healthcare providers to pool data from multiple systems, along with meta data, into a massive centrally accessible "mainframe" style system that can then be used for analytics and answer questions about care.

Eric Siegel, former Columbia University professor and author of best-selling book, Predictive Analytics, The Power to predict Who Will Click, Buy, Lie or Die, believes that big data is going to be as much a game changer in healthcare as in business, even if it is taking hold a bit later."It will make operations and clinical practice more effective and ultimately decrease cost,” says the author.

In response to concerns that there are not many takers for big data in health care, he shares success stories of universities and public companies. “Stanford University derived with predictive modelling an innovative method that diagnoses breast cancer better than human doctors in part by considering a greater number of factors in a tissue sample,” notes Siegel.

What about the accuracy of the data analytics prediction? According to Siegel there are several case studies which show that it has improved the accuracy of predicting disease progression from 70 to 78 percent. He also picks up the example of FICO, an American public company, which provides analytics and decision making services, which is also one of the case studies enlisted in his book. "It predicts patient compliance to drug prescriptions, identifying groups that will miss, on average, hundreds of days of regimen per year. Non adherence to medical prescriptions causes an estimated 125,000 premature deaths and over $290 billion in avoidable costs annually in United States alone,” says Siegel.

But what constitutes a sound data strategy? Tony Hussain, a Washington DC based healthcare strategist explains that one of the best ways that an organisation can embark on this journey is by creating a plan that aligns corporate strategy with data strategy."Companies should identify and clearly define what business insight (analytics) is desperately needed and select appropriate tools, balancing speed, cost and acceptance," adds Hussain.

Even though data science does make economic sense to management, there are several challenges involved in managing the data warehouse of an enterprise. Plus, one also has to keep in mind that 80 percent of data in healthcare is unstructured. According to Hussain, the biggest challenge is mobilizing the hospital management, all the way from the Board members, the CEO, CMO, the 'C' suite and middle management. "Big Data analytics transform the business landscape and place multiple new demands on top management teams, which in many case just do not have the bandwidth to completely understand and intelligently respond to those demands," says Hussain, who works with a prominent IT company.

Data is goldmine

Data analysts also point out that majority of healthcare companies do not know have a data-driven culture and do not treat data as an asset like money. Shahid N Shah, who has written extensively about healthcare IT, says that hospitals know how to treat their other assets like equipment, money, real estate, etc. better than they know how to treat data. “Data is left in the hands of IT departments without the business side really spending strategic time on how to really manage it as a valuable asset. Providers should be looking a proper Information Life cycle Management (ILM) approach and centralizing data in common pools so that each department doesn't manage it separately. Just like each department doesn't manage its own money without central budgeting and finance helping control it, data shouldn't be left in separate silos and applications to be managed on their own. So the challenge is not big data tools but ‘data as an asset’ thinking,” says the expert.

Closer home, in India, researchers of Public Health Foundation of India has developed apps to tackle grey spots like sanitation and hygiene. Raghav Sehgal, CEO of RxVault.in, the official  Electronic Health Records partner of Punjab Cancer Drive conducted by Punjab government and Max Healthcare, explains that data analytics helped in identification of problematic areas in Punjab and focused medical action in those areas. “We now know, by statistically significant datasets, the widespread of cancer in Punjab in all districts. As a part of the drive, we collected data of over 15,000 cancer patients,” says Sehgal. He also points out that collection of data is one of the biggest problems in Indian healthcare. "There is no data, structured or unstructured, in India for population. Few organized hospital chains keep records in electronic form, but most of them are limited to IPD patients and not OPD patients. Of course, that data is also limited to their hospital units. In order to truly benefit from big data, one needs to analyse data at individual patient level, as well as macro population level. Right now, no data is collected and digitized to that extent in India," he adds. "Just by using data analytics, hospitals, clinics and pharmacies can manage their inventory more efficiently and cost effectively, based on regional health trends. However, we can't do this, without collection of data," adds the Sehgal.

Size doesn’t matter

 But not everyone is enamoured by this technological revolution. There are critics like Vince Kuraitis, a healthcare consultant and author of e-Care Management blog, who calls for a cautious approach. He asserted in The Healthcare Blog that small data approach might be good enough for medical groups, while being more immediately implementable, and lot less costly. "We are not convinced in other words that the problem for Accountable Care Organisations is a scarcity of data or second rate analytics. Rather the problem is that we are not taking advantage of, and using more intelligently, the data and analytics already in place, or nearly in place," states the website.

Shahid N Shah is also of the opinion that big data is a bit over hyped presently. “But when genetics and proteomics become main stream then we'll need massive data technologies because genome and proteome data is even bigger than the kinds of big data in use today," says Shah. The analytics required on existing data sources of healthcare transactions such as claims, HR, clinical (except genomic), images, digital chemistry, and related information is manageable using traditional data tools. The issue in healthcare is that most of digital data available remains unstructured and therefore more difficult to analyze," notes Shah.
According to Nasscom study published in 2012, shortage of data scientists is one of the challenges that organizations have to face to implement big data."The big data phenomenon has led to an increasing demand for data scientists.-professionals conversant with both business context and data analytics- who play a crucial role in extracting insights from large data sets," states the study of the industry body.

Despite these challenges, senior healthcare executives are quite optimistic about adopting Big Data. According to them, the three major areas that big data would create an impact are evidence based medicine, telemedicine and clinical research.  “We have digitized our medical records and it is about 60 Terabytes now. However, the rate of growth of data is expected to grow at a much faster rate. We have been using data analytics to support our marketing programmes and to reach out to customers,” says Jagannath MS, Chief Financial Officer of Columbia Asia Hospitals. “As our data grows we are planning to look at trends based on data and images to enhance the quality of our medical decisions,” concludes the CFO.

The article appeared in the anniversary issue of Healthcare Executive.