The South African healthcare industry is in tremendous flux. Costs are spiralling out of control, medical aid customers are dwindling, fraud is a major problem, tremendous consolidation is leaving a mere handful of players, and HIV/Aids is having a significant impact.
How should healthcare providers, administrators and medical aid schemes be managing all this?
"The major issues for these organisations are risks and costs," says Corey Springett, business development manager at SAS Institute, the market leader in providing a new generation of business intelligence software and services that create true enterprise intelligence. "Analytical tools, such as predictive modelling, data mining and neural networks, are suddenly an absolute necessity for the industry."
These tools are able to identify unusual patterns, answer 'what if' questions, and analyse the generous volumes of data generated by the healthcare industry. They therefore help understand and manage risk, identify patients heading for high-risk conditions, reduce fraud, help with accurate pricing and provide intelligent decision support.
"Conventional wisdom no longer applies in many areas of healthcare," says Springett. "The traditional paradigm that the larger the customer base, the healthier the medical scheme, no longer holds true. It is now accepted that a large customer base is only healthy if those customers are profitable.
"Before the advent of HIV/Aids, younger medical aid members 'carried' the old," he says. "This has now been reversed and the old are having to 'carry' the young. Risk profiles have changed radically, and healthcare providers need to understand these."
While hospitals and other healthcare providers try to make a living in difficult times, the big drive for administrators and schemes is to cut provider costs. "Providers are having to analyse historic data to better manage their patient base," says Springett. "They need to identify risk, and then identify the most appropriate vehicle to manage that risk."
Understanding risk involves identifying patterns within patient behaviours that give early warning signals about potential problems, and much work has already been done with regard to diseases such as breast cancer, diabetes and hypertension.
"Once individuals have been profiled as likely to suffer from a certain disease, they can be put into managed care programmes," says Springett.
American Healthways, the largest care enhancement company in the US, for example, is using SAS predictive modelling to help identify patients who are trending toward a high-risk condition.
"This gives our nurse care co-ordinators a head start in identifying high-risk individuals. Then we can take steps to improve their quality of healthcare and prevent health problems in the future," explains Mark Ridinger, MD, the company's senior vice-president and chief science officer.
"Business intelligence software will never replace doctors," says Springett. "However, while the human brain can distinguish between five and 15 variables, data mining examines thousands of variables at the same time, and understands the inter-relationship between them."
Technology can assist by examining the results of numerous tests, scans, x-rays, examinations and procedures, and then predict an outcome.
Sophisticated business intelligence software can also be used to monitor the effect of healthcare programmes for certain individuals, and to answer questions such as which programmes are more effective for which individuals. Providers may also want to profile the effectiveness of particular treatments for certain conditions. They would look for patterns that indicate that treating a particular disease with a certain drug has a consistently high success rate, for example.
Analytic solutions can also be used to focus upon certain individuals, such as to examine the risk of a particular patient going into hospital to have a hip operation. Analysts could determine the chances of complications developing. In order to contain costs, it is vital that medical schemes understand risk. Their customer bases are dwindling as many people can no longer afford healthcare. They therefore need to analyse their customers, identify the good risk ones, put strategies in place to retain these, and adjust pricing accordingly. They also need analytics to help them understand which procedures they can push the providers on for discounts, and which can be grouped for fixed fee or capitation.
Administrators may want to identify risks among healthcare providers. They may want to identify doctors who are more expensive than most others, for example those who tend to write out far more prescriptions than the norm.
Another major issue facing the healthcare industry is fraud. Syndicates, often involving a provider and administrator, are defrauding medical schemes of millions each year.
"This kind of fraud is often enormously difficult to detect without using analytical tools," says Springett. "SAS solutions, for example, can be used to identify unusual patterns in prescriptions, or combinations of drugs that are subtly different to the norm. These often indicate fraudulent practices."
Utah's Bureau of Medicaid Fraud uses SAS to analyse Medicaid billing information and point investigators toward doctors whose billing practices may indicate fraud.
Prior to the introduction of SAS software, the Bureau of Medicaid Fraud had to sift through hundreds of pages of computer printouts to detect billing fraud. Because this method was so cumbersome, very little fraud was uncovered.
"Not only has our conviction rate increased, but our case load has increased along with it. We're simply able to handle more work now that we have SAS to help us," says Dr Terry Allen, statistician for the bureau.
Business intelligence can also be used for new product development. For example, if a medical aid scheme identifies a group of clients who visit the GP hundreds of times each year, it can then design a product or programme specifically for these people.
Pharmaceutical companies also make extensive use of SAS software for new product development.
"SAS software allows us to translate enormous amounts of raw data into knowledge. This is the basis for our continued efforts for innovation in the drug discovery research process," says Michael Engels, principal scientist, Janssen Research Foundation.
Springett says that the major strength of SAS is that it can get to data no matter where it is and then transform it into information. Finally, it turns the information into knowledge, giving the company sustainable competitive advantage.
Share