Rationale for big data analytics in healthcare is cogent, however implementation is one of the biggest hurdles providers face along with data security. To ride this new wave of business intelligence, Indian healthcare providers must realise that big data is a necessity and not a luxury By M Neelam Kachhap
If you have a social media account, chances are that you would have encountered a wishful debate/ discussion on big data analytics in recent times. Big data is the current buzzword and by all means, it is going to affect healthcare. But if you are not from the IT domain it’s difficult to gauge and keep track of the conversation on big data. In the present healthcare business environment, providers need to understand that big data analytics is a necessity and not a luxury and so is the understanding of big data analytics.
Big data analytics has enormous potential to impact healthcare positively by improving quality of care, saving lives and lowering costs. “Fundamentally, big data is helping organisations become more productive, efficient and reduce costs. Like many other industries, healthcare has adapted to data analytics not only for its financial returns but also for improving patients’ quality of life,” says Arvind Sivaramakrishnan, CIO, Apollo Hospital Enterprise, Chennai.
So, what is big data?
Professor Wullianallur Raghupathi from Fordham Graduate School of Business New York, US describes big data in healthcare as ‘electronic health data sets so large and complex that they are difficult to manage with traditional software or hardware; nor can they be easily managed with traditional or common data management tools and methods.’
For some time now, Indian providers have been using electronic health records (EHR) and hospital information systems (HIS) to make their organisation productive and profitable. These technologies collectively generate a lot of data. “Indian healthcare industry is engaged in generating zettabytes (1021 gigabytes) of data every day by capturing patient care records, prescriptions, diagnostic tests, insurance claims, equipment generated data for monitoring vital signs and most importantly the medical research. Growth of the digital data would be exponential and explosive in the next two years,” explains Niranjan Ramakrishnan, CIO, Sir Ganga Ram Hospital, Delhi. According to a industry report, California-based managed care consortium Kaiser Permanente is believed to have between 26.5 – 44 petabytes (1,000,000 gigabytes) of potentially rich data from EHRs.
“Organisation like us have data close to 500 terabyte of information,” informs Ashokkan VRS, Group CIO, Columbia Asia Group. “Healthcare as an industry should definitely have data in exabytes,” he adds.
However, ‘big’ in big data analytics not only defines the size but also the quality and complexity of data. “Big data could be defined as the total comprehensive data about an entity encompassing all sources. To better understand big data, it is important to understand what data is and how it differs from information, which are quite often thought of as one and the same. Data should be considered as raw information with or without filter, duplication, or structure that forms the building block for information. Transformation of data to information happens when one adds some logic to present a particular fact or view point. Information gathered from big data is often more substantial and unique, hence its value,” explains Sumit Singh, CIO, Wockhardt Hospitals, Mumbai.
Big data is also defined as large volumes of high velocity, complex and variable data that require advanced techniques and technologies to enable the capture, storage, distribution, management and analysis of information.
“Big data is a term which describes the exponential growth and availability of structured and unstructured data,” says Ravi Ramaswamy, Sr Director & Head – Healthcare, Philips Innovation Campus. He further describes the characteristics of big data by the 4Vs: volume, velocity, veracity and variety. (As described by the research and advisory firm Gartner)
Volume: It’s the quantity of data which gets generated and denoted in peta/ exa/ zeta bytes of data.
Velocity: Data is streaming in at an unprecedented speed and must be dealt with in a timely manner. RFID tags, sensors and smart metering are driving the need to deal with torrents of data in near-real time.
Variety: Data today comes in all types of formats. Structured, numeric data in traditional databases. Information created from line-of-business applications. Unstructured text documents, email, video, audio, stock ticker data and financial transactions. Managing, merging and governing different varieties of data is something many organisations still grapple with.
Veracity: In addition to increasing velocities and varieties of data, data flows can be highly inconsistent with periodic peaks. Is something trending on social media? Daily, seasonal and event-triggered peak data loads can be challenging to manage. Even more so with unstructured data involved.
“As per SAS Institute, a fifth element is also to be considered, which is complexity,” says Ramaswamy. “Today’s data comes from multiple sources. And it is still an undertaking to link, match, cleanse and transform data across systems. However, it is necessary to connect and correlate relationships, hierarchies and multiple data linkages or your data can quickly spiral out of control,” he explains.
Sources and ownership of big data
In the US and other developed countries, national registries and state departments collect health related data and aggregate it over the years. Thus making big data available for scientist to work on. However, in India health records are aggregated and stored by individual health organisations. There is a fair chance that this data may be in duplicates and is difficult to access. Having said that, some health organisations, for the larger benefit of the patients, may agree to share the data but even then it is a herculean task to get all similar data on one platform.
Scope for big data analytics in India
According to a report ‘Big Data Vendor Revenue and Market Forecast 2011-2026’ by Wikibon the US Big Data market reached $27.36 billion in 2014 and is slated to grow to $84 billion in 2026 . According to the report, one of the factors driving growth of the big data market was the increasing establishment of big data-driven decision making as a key strategic priority in board rooms and C-suites across vertical markets but particularly in the financial services, retail, healthcare and telecommunications industries.
Use of Big Data in Healthcare
Clinical Operations
Comparative effectiveness research to determine more clinically relevant and cost-effective ways to diagnose and treat patients
Research & Development
- Predictive modelling to lower attrition and produce a leaner, faster, more targeted R&D pipeline in drugs and devices
- Statistical tools and algorithms to improve clinical trial design and patient recruitment to better match treatments to individual patients, thus reducing trial failures and speeding new treatments to market
- Analysing clinical trials and patient records to identify follow-on indications and discover adverse effects before products reach the market
Public Health
- Analysing disease patterns and tracking disease outbreaks and transmission to improve public health surveillance and speed response
- Faster development of more accurately targeted vaccines, e.g., choosing the annual influenza strains
- Turning large amounts of data into actionable information that can be used to identify needs, provide services, and predict and prevent crises, especially for the benefit of populations
Evidence-based Medicine
- Combine and analyse a variety of structured and unstructured data-EMRs, financial and operational data, clinical data, and genomic data to match treatments with outcomes, predict patients at risk for disease or readmission and provide more efficient care;
Genomic analytics: Execute gene sequencing more efficiently and cost effectively and make genomic analysis a part of the regular medical care decision process and the growing patient medical record
Device/ remote monitoring: Capture and analyse in real-time large volumes of fast-moving data from in-hospital and in-home devices, for safety monitoring and adverse event prediction;
Patient profile analytics: Apply advanced analytics to patient profiles (e.g., segmentation and predictive modeling) to identify individuals who would benefit from proactive care or lifestyle changes, for example, those patients at risk of developing a specific disease (e.g., diabetes) who would benefit from preventive care
On the other hand, Indian business intelligence (BI) software revenue is forecast to reach $150 million in 2015, a 15 per cent increase over 2014 revenue of $133.8 million, according to Gartner. “Anecdotal studies indicate that the Indian healthcare sector is expected to contribute around 12 per cent of the big data generated in India. It is expected that this number will grow to 25 per cent of the overall data generated by 2017,” predicts Ramaswamy.
Advantages
Big data analytics generate actionable insights which can be used to predict disease outcomes, plan treatment protocols and for strategic organisational planning. By digitising, combining and effectively using big data, healthcare organisations ranging from single doctors practice to small and large hospitals to national hospital networks stand to benefit.
“Data analytics can help hospitals in financial planning, supply chain management, humane resource management and quality care delivery,” says Sivaramakrishnan. “Decrease in re-admission rates, predictive algorithms for diagnostics, real-time monitoring of ICU vacancies are some of the practical applications of big data in hospitals,” he adds.
Big Data in Healthcare
- Clinical data: Doctor’s notes, prescriptions, machine generated data, large format of images, cine sequences, scanned documents which are generated during clinical care and are not analysed as normal text data analysis
- Genomic data: Data acquired from gene analysis and sequencing
- Health tracker data: Data acquired from various devices, sensors, home monitoring and telehealth
- Web and social media: Health related data tweets, posts and publishers
- Health publications and clinical reference data: Clinical research, drug information, disease information, ministry and health body reports
- Other releated data: Personal preferences, behaviours etc. Administrative, commercial, socio economic, population etc.
According to Prof Raghupathi, potential for big data analytics in healthcare to lead to better outcomes exists across many scenarios. For example; applying advanced analytics to patient profiles to proactively identify individuals who would benefit from preventive care or lifestyle change; or creating new revenue streams by aggregating and synthesising patient clinical records to provide data and service to third parties like licensing data to assist pharma companies in identifying patients for inclusion in clinical trials.
In the public health domain, big data helps to analyse disease patterns to improve public health surveillance and speed response. “Big data analysis can help public health department to understand disease trends, help control sudden breakouts, lastly, helps in building awareness with facts and data,” says Ashokkan.
Cases in point
There are examples across the globe where big data analytics has benefitted healthcare organisations. The Institute for Health Technology Transformation, US cites a famous example of Kaiser Permanente which associated clinical data with cost data to generate a key data set, the analytics of which led to the discovery of adverse drug effects and subsequent withdrawal of Vioxx from the market in US. An IBM report cites the example of North York General Hospital, a 450-bed community teaching hospital in Toronto, Canada, which uses real-time analytics to improve patient outcomes and gain greater insight into the operations of healthcare delivery. North York is reported to have implemented a scalable real-time analytics application to provide multiple perspectives, including clinical, administrative, and financial. The Rizzoli Orthopedic Institute in Bologna, Italy, is reportedly using advanced analytics to gain a more ‘granular understanding’ of the clinical variations within families whereby individual patients display extreme differences in the severity of their symptoms. This insight is reported to have reduced annual hospitalisations by 30 per cent and the number of imaging tests by 60 per cent. In the long-term, the Institute expects to gain insight into the role of genetic factors to develop treatments.
The Hospital for Sick Children (Sick Kids) in Toronto is using analytics to improve the outcomes for infants prone to life-threatening nosocomial infections. It is reported that Sick Kids applies advanced analytics to vital-sign data gathered from bedside monitoring devices to identify potential signs infection as early as 24 hours prior to previous methods.
Back home, Sivaramakrishnan describes infection control using data analytics at Apollo Hospital.
“Microbiology department has important roles to play in any potential outbreak situation, including early recognition of possible clusters and outbreaks, rapid notification of and collaboration with the infection control team, which requires maintenance of an organism bank,” says Sivaramakrishnan. “The microbiology laboratory should also act in a consultative capacity with the infection control team to help determine whether an outbreak is ‘real’ or a potential pseudo-outbreak due to contamination of specimens outside or within the laboratory,” he adds.
“The current process is laborious with manual statistical analysis by the Microbiology department and by the infection control team to get this output. Apollo Hospitals has developed in house analytical tools for use on our HIS. The system was created using Microsoft Business Intelligence tool and utilised Excel front end dashboard. Both Microbiology department and infection control team was granted access to the analytical tool. Since it’s an Excel front end dashboard training the staff was easy on the use of the analytical tool,” he further adds.
“Dengue seems to rear its ugly head in many part of the country regularly and effects many citizens and their families. With the availability of big data, many factors that help in its formation are identified and then alerts are sent ahead of time to prepare to handle the outbreak,” explains Singh.
Talking about examples of the benefits of big data analytics in India, Ashokkan says, “Very little is being done today, as the information technology adoption in healthcare industry needs a large revolution and standardisation in the Indian sub-continent.” Some examples of application in Columbia Asia are; determining the accuracy of diagnostic investigation reporting with cross match of data sets which vary from diagnostic images, culture reports, clinical notes, diagnosis identification etc. And customising healthcheck packages for customer segments,” he adds.
Sharing examples from Philips, Ramaswamy says, “A study over five years examined the impact of Philips’ remote intensive care unit (eICU) programme on nearly 120,000 critical care patients. The programme enables healthcare professionals from a centralised eICU centre to provide round-the-clock care for critically ill patients using bi-directional audio/ video technology, and a clinical decision support system. The study found that eICU patients, compared to patients receiving usual ICU care, were 26 per cent more likely to survive the ICU, and were discharged from the ICU 20 per cent faster.”
Philips is now extending these initiatives by building an open digital platform that can link to all kinds of devices, allow doctors to feed information about patients, allow patients, relatives and doctors to be connected to each other, and do large scale analytics. “Any doctor anywhere will be able to look into the entire history of a patient to do better diagnosis. Relatives and professional care folks can get immediate alerts if something goes wrong. And the vast amounts of data collected on the platform can lead to algorithms that can improve diagnoses, figure out what works for what kind of patient,” explains Ramaswamy.
Challenges
According to Prof Raghupathi, “Healthcare data is rarely standardised, often fragmented, or generated in legacy IT systems with incompatible formats.” This is one of the biggest challenges in India. “Over a period of time, India will build a staggering amount of healthcare data but it would be spread among hospitals, primary care providers, researchers, health insurers, and state and central governments—just to name a few. Each of these act as a silo, preventing data transparency across the healthcare system,” says Ramaswamy.
Another challenge would be veracity of data. “Different types of data from different systems, adherence to standard formats, inter-operability issues and homogeneity would also pose a great challenge,” says Ramaswamy. In addition to aggregating a massive amount of data, there’s the challenge of maintaining patient privacy. According to Dwayne Spradlin, CEO of the non-profit Health Data Consortium, private healthcare data is critical to big data’s success, it doesn’t mean that private data will become public. Figuring out how to leverage that information to deliver better quality care to patients while keeping it secure is a major challenge.
Policies related to privacy, security, intellectual property, and even liability will need to be addressed in a big data world. Organisations need to not only put the right talent and technology in place but also structure workflows and incentives to optimise the use of big data.
Conclusion
Healthcare in India is witnessing a new wave of competition with foreign investments and disruptive technology and this will further intensify as the turf gets structured. The organisations that are looking at big data now are the ones with lowest-hanging fruit, and their success stories will help other providers see how they can make their own ventures, fruitful.
Reference: Raghupathi, W., Raghupathi, V.: Big data analytics in healthcare: promise and potential. Health Inf. Sci. Syst. 2, 1, 3 (2014).