5 Ways Data Analytics is Already Revolutionizing Modern Healthcare

Ilan Hertz
Ilan Hertz  |  June 1, 2018

While the human side of healthcare is often the most visible, technology plays a huge role.

Data analytics is part of that technology, helping healthcare organizations to save lives, as well as time and money. It’s one of the few solutions for effectively understanding and leveraging the petabytes of data available from legacy systems, medical health records, and public databases.

Much of this data is gathered for regulatory reporting and requirements. With healthcare analytics, however, the data can also be put to further use to improve care, cure, and even prevent illness.

Spotting patient condition changes in intensive care units

When it comes to monitoring a patient’s vital signs and sounding the alarm when danger looms, data analytics have a big advantage. They are always on. No change in a patient’s condition will escape them, day or night. Better still, data analytics can spot downtrends early in cases where human caregivers are less able or too busy to predict and prevent crises. 

For example, the condition of child patients in the cardiac ICU of the Boston Children’s Hospital in the U.S. can change suddenly and significantly. Data analytics, and more specifically bedside predictive data analytics, help keep young patients alive by bringing together data from different bedside monitors in a single at-a-glance ‘stability index’ to show each patient’s condition overall.

Data analytics is also used in hospitals at the time of patient admission. It helps staff assess risk and select the treatment that is best aligned with the patient’s condition and profile. With additional input from national guidelines, data analytics and data visualization through healthcare dashboards can provide reliable, easy-to-understand risk-level indication for operations or other caregiving actions.

Precision and evidence-based medicine

Picking out insights about specific illnesses from healthcare data and zooming in on probable causes are two further strengths of data analytics. At a macro level, data analytics boosts efforts in the fight against cancer and the development of preventative measure against heart disease and diabetes.  

At a micro or individual patient level, it gets results faster than the cumbersome checklist diagnosis in which a physician must plow through lists of points to check a patient’s condition.

At the Icahn School of Medicine at Mount Sinai in New York, data analytics is helping staff see the crucial difference between two otherwise very similar heart conditions.

While the final call is still down to a human physician, the data analytics algorithms process and analyze large quantities of cardiac ultrasound data to help doctors hone in on the right diagnosis. By making the data analytics program and its healthcare dashboards available remotely, smarter heart medicine could then be made available to even the most resource-constrained health centers.  

Natural language processing (NLP)

Data analytics are powerful number crunching tools. However, interactions in healthcare between people don’t happen in binary code. They take place using natural language, like the words on this page. Natural language processing, a blend of data analytics and machine learning (part of artificial intelligence), lets a computer make sense of natural language. This also means that the power and speed of computer systems can be used to tackle the mountains of natural language notes in patients’ health records, as well as from patient satisfaction surveys and recordings of interviews and operations.

Intermountain Healthcare in Salt Lake City, U.S., has found natural language processing to be invaluable in identifying heart failure patients visiting the hospital for other reasons such as surgery or respiratory problems. The hospital uses its system to read data from 25 types of free text documents within electronic health records to make sure patients and cardiovascular experts are alerted in a timely way to potential problems.

Reducing readmissions

Ideally, a patient leaving a hospital should never need to go back. However, readmissions can happen. It is then in everybody’s interest – patient, caregiver staff, healthcare insurer – to keep them to a minimum. The Parkland Health and Hospital System in Dallas, U.S., has developed a data analytics algorithm to predict the risk of readmission of patients with heart failure. Patients considered to be at high risk receive education, phone support, and follow-up appointments. 

Similarly, hospitals can use the detailed data from health records and data analytics to spot patients likely to need readmission. By identifying this factor early in the patients’ initial hospitalization, the staff can organize care to prevent the need to return where possible. Data analytics also allow performance of healthcare institutions to be assessed over time. This makes value-based reimbursements more accurate and fairer, encouraging the institutions to reduce readmissions yet further.

Population health management 

Targeting prevention as well as cure, healthcare establishments can use data analytics with data sources like Google Maps and free public health data to visualize hot spots for population growth or concentrations of specific diseases. They can then adapt their skillsets, resources, and services in a strategic plan to manage these changing situations.

The Veteran’s Health Administration (VHA) in the U.S. is using data that it has collected from its patients over the past three-plus decades. Data analytics allows it to predict overall outcomes for its constituents such as the risk of hospitalization. The VHA started building a data warehouse for the patient data about 10 years ago. The investment was sizeable, and the project continues to claim about 5% of the total health budget of the VHA, but it also delivers a net return measured in billions of dollars.

Healthcare data analytics accessible to all

New approaches to data analytics and the systems that run them mean that healthcare establishments can get the advantages without needing to hire in an army of data scientists. Non-data specialists can explore data through intuitive healthcare dashboards and ask ad hoc questions as they like. This is another reason why data analytics is being used today in many practical healthcare applications. 

The success of the establishments and healthcare analytics described above is therefore within the grasp of many healthcare organizations, allowing them to optimize the use of their resources and maximize the benefits to their patients.

Ilan Hertz
Author

Ilan Hertz

Ilan Hertz is Head of Lead Generation at Sisense, the leader in simplifying business intelligence for complex data, offering a powerful business intelligence software. Ilan uses his domain expertise and data-driven methodologies to lead digital marketing efforts at Sisense.