How Data Analytics Is Reducing Transfusion Use and Length of Hospital Stays

A crime-fighting mathematics genius once expounded: “Everything is numbers,” and while the source may be a work of fiction, the statement is quite true. Virtually anything can be summed up using statistics that can help predict future events and conditions.

Within healthcare, the feeling today is that to continue treating patients over many years without tracking the trends of their outcomes would border on malpractice.

With the advent of electronic health records, it is now possible to access larger, more comprehensive volumes of statistics than ever before, relating to medical practices, hospital stays and patient outcomes

New techniques are revolutionizing transfusions of both whole blood and blood products. According to a study conducted by the Mayo Clinic, hospitals can cut back on blood transfusions that are not specifically needed by adopting comprehensive blood management programs for their patients.

Patient blood management involves minimizing blood loss during surgery and returning or recycling the patient’s own blood collected either before or during surgery. According to the findings, blood management programs, arrived at with the help of thorough data analytics, can also reduce the length of hospital stays and improve patient outcomes.

This study, running from 2010 to 2018, involved 400,998 patients at a pair of Mayo Clinic Hospitals and found that with appropriate patient blood management programs, a third of all allogeneic transfusions could be eliminated, and hospital stays shortened up to 15%. There was also a noticeable reduction in the number of critical in-hospital events, like strokes, myocardial infarctions, venous thrombosis or urgent respiratory issues.

The potential savings to the two institutions were estimated at about $7 million annually.

The ability to reduce or even eliminate some allogeneic transfusions can be beneficial to the US national blood supply. An over-reliance on transfusions frequently results in adverse patient outcomes, so it is good to get the use of them under control.

It’s only through repeated observation of the dismal outcomes experienced in crowded hospital wards 100 years ago that transmission of pathogens and subsequent infection control protocols were deduced. Fast forward a century, and we find that statistics alerted us to the burgeoning opioid abuse crisis that had already established a firm foothold in the western world.

Clinical studies are conducted by capturing data on every aspect of a patient’s pre-treatment, clinical actions taken during treatment and a careful post-treatment monitoring period. Now that we have a thorough picture of ongoing hospital activities –both the view from 30,000 feet and the granular data on each medication dispensed and vital signs observed – we can begin to analyze this data to quantify patient outcomes and qualify the effectiveness of treatments.

Analyzing data can begin well before the patient contracts any malady. Factors such as lifestyle, diet and hygienic habits over the patient’s lifespan can inform treatment and outcomes at almost any stage of life. In fact, much can be learned from these statistics to develop new modalities of preventative healthcare.

The war on major diseases such as cancer is being fought largely with data analysis. Close scrutiny of statistics has revealed geographic hot spots and linked disease patterns to pollution sources. Researchers from institutions around the globe can share data on specific tissue samples and the records of the patient who produced them.

Since the path to new treatments and improved pharmaceuticals is a long, time-intensive one, using data analysis can help new drug developers determine what the next widespread ailment will be. The medical and pharmaceutical industries invest tremendous capital into ramping up for the next wave of illness, so they’ll be very interested in compressing that timeline with predictive capabilities.

Patients themselves are becoming more engaged in their own healthcare practices. New generations of wearable digital monitors can keep comprehensive records of a patient’s vital signs and symptoms, thus tracking health histories which can point a predictive finger toward potential critical conditions to come.

Telemedicine has already proved itself on the medical battlefield, and data plays a large part in remote healthcare. Algorithms in medical imaging can help immensely with diagnosis, and ready files of millions of other cases help telepractitioners compare and arrive at a much-needed and prompt diagnosis.

The recent advancements in EHR and the introduction of automation and “smart systems” into almost every facet of modern healthcare have come a long way to bringing healthcare closer to the exact science we all know it can be. 

However, there has been some resistance and blocking of access to assets that are needed to make this machine function at its peak.

These factors include:

• Incompatible systems of acquiring and organizing data.
• Issues of patient confidentiality, as laws governing confidentiality can differ widely from one jurisdiction to another. 
• Some institutions are reluctant to allow making data they see as proprietary available to fellow researchers.

Supporting an all-inclusive database of EHR content will serve researchers, patients and the healthcare industry well into the future. This aspect of technology use in healthcare has vast benefits for the health of our nation.

About the author: Chris Patrick is the CEO of Nuvodia based in Spokane, Washington. For the last 20 years, Nuvodia has helped customers by providing a private cloud, managing their infrastructure, and assisting with multi-cloud strategies and implementations.