Better data, better diagnoses, less fraud
By Adrian Smolski, Solutions Architect ANZ, MapR Technologies
Wednesday, 22 February, 2017
Australia’s health sector is sitting on a data goldmine, the full exploitation of which could lead to better health outcomes and reduced costs.
The Australian healthcare sector is undergoing continuous automation and digitisation, with many new initiatives aiming to make it easier for patients to use medical services and manage their health, while reducing costs.
In the past few months alone, Cairns Hospital became the largest regional hospital in Australia to go paperless through its Digital Hospital initiative, while eHealth NSW rolled out a major digital transformation project across its electronic media records in NSW hospitals. While these are very promising developments, the sector is still going through tough times, trying to find answers to rising costs that are driving patients to cancel their private health insurance.
Digital analytics technologies can help healthcare organisations stay on top of these and other challenges. Let’s pick just two — the delivery of better medical diagnoses, and fraud management.
Tapping into big data
In recent years, medical institutions, government entities and hospitals have collected massive quantities of data that are yet to be effectively utilised to benefit the industry and its customers. Australia’s health sector is sitting on a data goldmine, which could be the key to solving critical issues that are costing our economy and our citizens’ health.
And still, many healthcare professionals and organisations have yet to invest in sophisticated data analytics solutions to help uncover opportunities hidden in unstructured data (80% of healthcare information) that comes from a wide range of sources — professional and personal medical devices (eg, health apps and wearables), doctors’ notes, lab results, correspondence between professionals and institutions, hospital and clinical data, and financial data.
Medical diagnoses are now expected to be fast and precise, with healthcare practitioners and facilities looking to provide more proactive and simplified care for their patients. In today’s digital age, in-hospital patients’ vital signs are continuously monitored and every individual is empowered to monitor their own health at anytime and anywhere through wearables, apps and smart devices.
These sources of data represent opportunities to fine-tune diagnostics, either by preventing a condition or by better addressing it. All the data collected by medical monitors, whether they are in-hospital or worn by the patient in their everyday life, needs to be analysed, cross-referenced and effectively leveraged to affect results. Combining real-time event data with machine learning algorithms can provide physicians with insights to enable lifesaving decisions and effective interventions.
For example, Flinders University recently developed an analytics program that is able to simplify dysphagia diagnoses, speeding up results and eliminating the need for X-rays. Another example — predictive modelling of data derived from electronic health records (EHRs) is being used for early diagnosis, for example for reducing mortality rates linked with congestive heart failure and sepsis.
Another significant benefit that effective data analytics can bring is the clear, accurate financial insight that can lead to fraud reduction and prevention.
In 2016, Medicare lost $1.6 million through fraud by doctors and others, and private health insurers have been shown to be still very much challenged by the fraud issue. Finding ways to prevent and monitor fraud would benefit the entire industry, as well as patients.
Dig data and analytics tools could be a game changer here, as the technologies available today can prevent, identify and neutralise fraud based on the analysis of unstructured data sets.
For example, based on patient records, billing details and history, healthcare organisations can use analytics to detect anomalies such as a hospital’s over-utilisation of services within short time periods, receipt of services from different hospitals in different locations simultaneously or identical prescriptions for the same patient filled in multiple locations. Based on historical data and specific patterns, analytics tools can also predict future risks and help prevent fraud before it happens.
In the United States, for example, the Centers for Medicare and Medicaid Services prevented more than $210.7 million in healthcare fraud in one year using predictive analytics. United Healthcare also transitioned to a predictive modelling environment based on a Hadoop big data platform, identifying inaccurate claims in a systematic and repeated fashion which generated a 2200% return on its big data and advanced technology investments.
As EHRs continue to grow and evolve, combining records with data analysis across multiple data sources will enable better diagnoses and services for patients, while reducing costs and providing better patient experiences.
In the wake of the weekend's WannaCry ransomware attack, Microsoft has renewed its call for...
Measurement of operational customer experience can assist in recovery from online failures.
The Productivity Commission's final report into open government data calls for a new...