Skip to main content

How is data analytics transforming healthcare outcomes?

(Image credit: Image Credit: Shutterstock/Sergey Nivens)

Data analytics presents tremendous opportunities in the right hands. It can help us to grow our businesses, make our lives more efficient, as well as help worthwhile causes. I find it particularly rewarding working with healthcare organisations, where data analytics can help to make a real difference to people’s lives. If you look past some unfortunate white elephants, technology has helped both public and private sector immensely in coping with the increasing demands of our society’s healthcare needs. This is thanks to the significant advancement in the digitisation of patient information, wearable technologies and remote monitoring, and more recently AI.

Data analytics has also played an integral part in the present digital revolution. In the near future, we will see an increasing number of organisations embracing the power of big data, allowing healthcare professionals to stay one step ahead of the game by anticipating needs and potential problems.

With its highly organised record-keeping practices, hospitals and clinics have been unknowingly preparing for data analytics for many years. Patient records are an enviable and extensive goldmine of useful information. For example, non-profit organisation Piedmont Healthcare comprises eight hospitals and has more than 555 billion data points to draw from. It employs these data to provide healthcare professionals with the means to drastically improve diagnosis, pharmaceuticals use and dosage. By tapping into insights based on historic information from a patient or patient group, it is easier for healthcare professionals to arrive at the right decisions and take the best action.

To date, much data analytics discussion in healthcare has been dominated by identifying efficiencies and reducing costs. However, putting aside these advantages, I believe there should be a more concerted effort to apply data analytics in daily healthcare operations to directly fight diseases with data analytics and help avoid preventable harm.

The complete elimination of diseases

Data analytics has the capability to help to completely eradicate life-threatening diseases such as malaria. Although the air-borne disease is still rampant throughout Zambia, with 5 million reported cases last year, this number could be reduced to zero by 2021. The Zambian government has plans to break this cycle forever by joining forces with PATH (Program for Appropriate Technology in Health) and eight technology sponsors under an initiative called Visualise No Malaria.

The insight gained from operational data analytics has given healthcare workers the ability to both track and predict the disease’s movements, which is crucial to halting its spread. By visualising field worker and weather data in maps, its operatives now intervene to prevent outbreak in areas at risk, where they could previously only react to cases of infection.

When outbreaks do occur, healthcare staff at Zambia’s National Malaria Elimination Centre use data analytics to make decisions on the optimal times and places to distribute life-saving resources such as disinfectant, medicine and bed-nets. This targeted approach is working well and things are looking optimistic. As they collect more data, healthcare staff will refine their analytics and gain a greater understanding of the disease so they are better placed to eradicate it.  Through a sustained data-driven strategy, we should see the end of malaria in Zambia in just three short years.

Avoiding unnecessary harm

There are many instances where harm such as treatment mistakes or post-operative infections which could be avoided. With the help of data, healthcare providers stand a better chance of granting a safer patient experience. Data analytics can produce answers in near real-time using the latest data. Piedmont Healthcare looked to data analytics to help, and after just one year achieved a 40 per cent reduction in unnecessary harm. This figure was measured by tracking 30 metrics including readmittance rates.

Infection Control

Hospital staff work extremely hard to keep the hospital environment as sterile as possible, but infections can – and do – break out. For vulnerable patients, infections not only spread quickly, but they can also be twice as deadly as well. Piedmont Healthcare addressed this challenge by developing an infection dashboard to control infections. When an infection breaks out, staff can use recent data to help determine the root cause of infection and identify its origin. By tracking information like this, it also enables staff to better avoid infections in the future, as they have a better understanding of how they are breaking out and can take steps to prevent them. As a result of the dashboard, a number of common hospital infections including MRSA were eliminated in a number of Piedmont hospitals last year.

Extracting the most value from your data set

Presently, most large businesses have integrated some pockets of data analytics, or are in the process of evaluating technologies. However, this does not guarantee that the most value is being extracted from the data. There are several key considerations that organisations should bear in mind to ensure the effective use of data.

Firstly, the relevant staff need to have access to the insights gained from the data. This not only includes high-level decision makers, but also other workers so they have the power to carry out data-led actions autonomously.

Another important consideration, which is doubly important in a clinical setting, is speed. This means ensuring that reports and visualisations are produced in as close to real-time as possible and from the most up-to-date data. In the instance of an infection breaking out for example, each second is valuable to preventing its spread as outbreaks can quickly cause serious harm.

The last, but arguably most important, measure is to ensure that staff (not just the data science or analyst team) know exactly how to use the data. This means understanding which questions to ask of the data in order to produce useful answers and enable staff to make the right decisions. To accommodate for this, it is highly recommended that hospitals invest in designing visualisations that are simple and clear so that data can be interpreted easily by those who need it. 

In a few short years data analytics will emerge from research labs and become commonplace in everyday healthcare, and when applied correctly, this can only mean great leaps in healthcare will be made and more lives can be saved.

Mathias Golombek, CTO, Exasol
Image Credit: Shutterstock/Sergey Nivens