Analytics can ease the challenges faced by SA's healthcare system and can transform service delivery.
This is according to Goran Dragosavac, solutions manager for information management at SAS Institute SA.
Analytics can deliver insights about illness treatments and potential cures, as well as control disease outbreaks, help run hospitals efficiently, control costs and allocate staff where they are most needed, he says.
According to Dragosavac, this is especially true for descriptive analytics - which uses data to focus on what went wrong and assessing why certain outcomes are less than expected. Similarly, explanatory analytics, which identifies patterns and predicts outcomes to avoid preventable events, can also be a handy healthcare tool, he says.
Two basic categories of patterns exist, says Dragosavac: "desirable" and "undesirable".
"In the healthcare industry, an example of 'desirable' patterns would be segments of the population with certain characteristics that happen to have below-average mortality rates, low HIV/Aids infection rates, quick recovery times, etc. 'Undesirable' patterns would be the opposite of these - segments of the population that have very high (above-average) mortality rates and similar," Dragosavac notes.
Once attributes of such segments are known through the process of pattern extraction, these patterns can be acted upon through fortifying, supporting and replicating desirable patterns and breaking down undesirable ones, he adds.
"That can be as simple as finding which hospitals are linked to segments of high infant mortality rates and then going in those hospitals with a specific set of interventions aimed at reducing mortality rates," says Dragosavac.
The use of analytics can also minimise medical errors, notes Dragosavac. "The key to decreasing these errors is to properly identify them, analyse their causes, and then change the system to prevent them happening in future."
Some adverse events are not preventable, explains Dragosavac, such as a patient experiencing a life-threatening allergic reaction to a drug when the person has no known allergies.
"But there are many other types of medical errors, including equipment failure, infections, blood transfusion-related injuries, misinterpretation of medical orders, incorrect medicines or prescriptions, surgical errors and lab report errors," he says.
"By identifying patterns in these, it is possible to do what is necessary to educate the people involved - patients and caregivers - to reduce their incidents."
Dragosavac further states that analytics solutions can be used to look at factors like mortality, morbidity, re-admittance rates, changes in birth and death rates, time to recovery, and patients' own perception of their care and progress.
However, there are still barriers to adopting analytic models in SA, he notes.
According to Dragosavac, there are many good statisticians employed within SA's healthcare sector (private and public) and there are powerful analytics technologies available, but these skills are deployed more for tactical and application purposes, instead of being used for wider strategic goals.
"As with any other industry, healthcare can derive a wealth of benefits from turning analytics to the mass of data the industry generates."
The only distinguishing factor for the healthcare industry is that the benefits realised can save lives and ensure wellbeing, so their need is far more pressing, concludes Dragosavac.
Share