Case Study: Cloud Analytics for an Assisted Reproductive Market
Imagine you are the CIO of a healthcare firm and a critical business decision is pending marketplace insights. Your data science skills are limited, your data is too big for your spreadsheet application, and your budget cannot afford a proprietary software solution. What do you do?
Businesses often have situational analytic needs that require access across multiple data sources. Healthcare institutions, like many businesses, are severely constrained by their inability to derive analytic value from data in a timely manner. Cloud-based analytics offer several advantages to healthcare institutions with limited resources and budgets looking to make timely business decisions based on insights derived from data sources.
Most healthcare organizations are faced with three (3) common issues pertaining to their adoption of healthcare analytics:
- They lack the data science skills necessary to integrate a deluge of incompatible data sources;
- They require results in a timely manner and lack the time necessary to research, acquire and deploy a custom solution architecture;
- Their focus is more on analytical results and less on how the results are obtained.
The Monash IVF Group (Monash), with over forty years of experience, has assisted in the reproductive needs for over 35,000 babies. Recently, Monash teamed up with IBM to tackle challenging business questions through a combination of data science skills, methodology, cloud computing, and the Jupyter Interactive Notebook to produce timely analytic driven solutions.
The first Assisted Reproductive Therapy (ART) baby in the world, Louise Brown, was born in 1978. In 2012 there were 13,500 live births as a result of ART in Australia and New Zealand. This is a result of improvement in techniques, widespread availability of services, the delay in starting families in the western world, and an increased acceptance of in vitro fertilization (IVF) treatment.
Medicare Australia is the government agency which funds most health procedures in Australia, either completely or partially. The Medicare Benefits Schedule (MBS) is a ledger of all the conditions which Medicare funds, the rules governing who can provide and who can receive the service and the fees paid for each. IVF services can only be provided to women under 46 years old.
The volume of IVF services supplied in Australia has grown reliably at about 4% annually for the 10 years up to 2013. In the 18 months prior to March 2015, growth in the volume of services has been at 0% or possibly slightly declining. There have been occasional similar drops in growth, but never so pronounced or so prolonged.
Monash needed to better understand the dynamics and variables associated with the growth of the market for IVF services in Australia. They wanted to know if demand had peaked for the regional industry and if the market for IVF services in Australia had reached an inflection point.
Today, there are many healthcare institutions adopting predictive analytics to improve various aspects of their business operations.
Dan Gisolfi, Client Innovation Avocate of IBM Cloud Emerging Technologies, deployed an agile CRISP-DM methodology to help Alan Pritchard, Chief Information Officer at Monash. Given the 14hr difference in their geographical locations, the pair leveraged an IBM SaaS offering based on the Jupyter Interactive Notebook to iteratively prepare, clean, explore, mine and visualize data.
“Our data science consulting approach is to pair technical skills with domain experts. Our clients seek more than the timely discovery of insights, they desire to quickly and easily communicate their findings through visualizations and narration while also having the option to share their results as reproducible research. We have found Jupyter Notebooks to be an indispensable asset in our tool belt for data science consulting engagements.” – Dan Gisolfi, Client Innovation Advocate, IBM Emerging Technologies
“IBM helped Monash IVF to gain a deeper understanding of the IVF market in Australia and to forecast how it is likely to change in the next 3- 5 years. This was a somewhat complicated exercise because it required the consolidation and extrapolation of data from different sources each associated with various time periods. We needed to overlay internal company data about procedures and pricing, which is up to date, with data from the Medical Benefits Schedule and Pharmaceutical Benefits Schedule, both of which have a variable amount of processing delay, and population data and forecasts from the Australian Bureau of Statistics, which is up to 5 years out of date.
Working with IBM we were quickly able to make sense of the various sources of data and build a model which allowed us to gain a view of the likely changes in the market for IVF services over the next 3 – 5 years. Understanding a specialized medical service in the Australian health funding model throws up many challenges for anybody without prior experience but IBM’s dynamic and collaborative process for building models means that progress can be made very quickly while the data analysts are gaining domain specific knowledge. From the very first meeting, we were able to start building our forecasting model with Monash IVF subject matter experts advising IBM’s data scientists on the structure of the data and the desired modeling outcomes. Within a few meetings the IBM team had developed a very strong understanding of the data and the underlying business questions, and were helping to drive the modeling process forward. All of this done with the added benefit of remote collaboration using IBM’s cloud service.” – Alan Pritchard, Monash Chief Information Officer