A physician’s perspective: Reducing HACs with data science
This blog is part of a two-part series featuring guest authors who are on the front lines of healthcare. Today’s post features Tom Louwers, MD, MPH, Associate Medical Director, Healthcare Insights at KenSci. Dr. Louwers sat down with us to explain how KenSci is preventing death and disability from hospital-acquired conditions with data science. Learn more about how KenSci Clinical Analytics is revolutionizing the health care industry on Microsoft AppSource.
Hospital-acquired conditions (HACs) are a troublesome reality of the healthcare system, and include maladies like sepsis, adverse drug events, surgical site infections, clinical errors, and more1. Medical errors alone kill 251,000 Americans every year, making them the third leading cause of death in the U.S.2 At any given moment, 7% of hospitalized patients in developed countries – and 10% in developing nations – are acquiring at least one healthcare associated infection3. The good news is that hospitals are getting better: The U.S. Department of Health and Human Services found a 17% decline from 2010 to 2014 in HACs, resulting in 87,000 fewer patient deaths in hospitals and $20 billion in care cost savings4. Unfortunately, there is still a long way to go – only 2.3% of U.S. hospitals achieved a five-star rating from the Centers for Medicare and Medicaid Services5, and 769 hospitals’ Medicare payments in fiscal year 2017 were cut for having high rates of HACs6.
Care providers must first do no harm, and the struggle to stop HACs is led by data science. Leading organizations are using advanced analytics to minimize clinical errors, intervene against HACs earlier at lower costs, and ultimately save lives. Let’s take a look at how data is transforming the patient’s journey through the healthcare system – drastically reducing the risk of HACs.
Gain a deeper understanding of patient health and risks
Physicians often have too much data and too little time. Let’s take Larry, a 65-year-old patient in need of a new hip. Before he comes in to the hospital for his hip replacement surgery, the risk of sepsis may be one of the last things on his doctor’s mind. But sepsis, a condition that occurs when the body’s attempts to fight an infection are overwhelmed, is the number one killer of hospital patients – occurring in over 1 million patients in the U.S. every year and causing more readmissions at higher costs than pneumonia, heart failure, or heart attack7. Larry’s age, his hypertension, the fact that he smokes, his history of hospital admissions for pneumonia and the flu all may signal a weakened immune system and a higher risk of sepsis. Healthcare providers normally only have the time and tools to evaluate a limited number of Larry’s risk factors – but data science changes the game.
Aggregating individual patient data sets from Larry’s EHR (Electronic Health Record), ADT (Admission, Discharge, and Transfer) logs, and wearable sensor with pre-built data analytics models that mine millions of records, Larry’s care provider determines that he is at a high risk of sepsis. Larry’s care team sees his predicted risk level and his modifiable and unmodifiable risks, and acts on those insights to reduce the likelihood Larry will acquire sepsis. Before Larry comes in for his hip surgery, his care team contacts him to ensure he’s taking his current hypertension medications and following his prescribed pre-surgery regime.
Minimize errors and enhance care
Too often, HACs are the result of simple clinical errors. Luckily for Larry, his care provider leverages data to prevent even basic mistakes. Algorithms comparing Larry’s facility and care team to other providers indicated that they faced a higher rate of infection and suggested poor hand-washing compliance may be the cause. Acting on those insights, Larry’s hospital instituted a new, more rigorous hand-washing and clinical hygiene regime, significantly lowering infection rates. Before and after Larry’s surgery, his whole care team follows a meticulous sanitization routine.
Larry is wheeled out of the operating room after a successful hip replacement surgery, but his care team’s work is far from over. Post-surgery, a patient’s vital signs are eerily similar to those of a patient with sepsis, as the patient’s body reacts to the trauma of the operation. Advanced analytics surface something incredibly subtle that even an expert care team might miss – Larry’s respiratory rate is remaining relatively elevated. Though Larry’s vitals look like they are in the normal range for a 65-year-old male like him, a historical analysis of Larry’s baseline vitals reveal a lower-than-average respiratory rate. As his respiratory rate and temperature start to increase at a near-indiscernible rate, predictive models provide concrete clinical decision support to Larry’s care team – suggesting antibiotics to stop the subtle onset of sepsis. With data science and predictive analytics, Larry’s care team is empowered to distinguish even the most minute vital sign changes over time and benefit from actionable insights – ultimately enhancing his care and preventing progression to a more severe condition.
Once Larry’s condition has stabilized, his care team prepares him for discharge, leveraging data models to predict readmission risks, offer an optimal discharge plan, and determine the best post-hospitalization care facility. Balancing Larry’s risks from being released too soon and staying in the hospital too long, advanced analytics help maximize Larry’s care while minimizing costs, HACs, and the likelihood of readmission.
Clinical Analytics dramatically reduce hospital-acquired conditions
KenSci’s Clinical Analytics solution, built on Microsoft Cloud Technology, improves health outcomes and reduces the total cost impact of HACs like sepsis. The solution identifies patterns and predicts patients who are at high risk of sepsis and other HACs, enabling care managers to engage high-risk patients and modify their risks to prevent infection, readmission and mortality. Plus, it integrates easily with existing EMR systems – so care managers and care providers aren’t slowed down by having to learn and use a separate interface. Try the solution today on Microsoft AppSource.
________________________________________________________________________________________________
1http://www.beckershospitalreview.com/hospital-management-administration/50-things-to-know-about-the-hospital-industry-2017.html
2https://soundcloud.com/bmjpodcasts/medical-errorthe-third-leading-cause-of-death-in-the-us
3http://www.who.int/gpsc/country_work/gpsc_ccisc_fact_sheet_en.pdf
4http://www.beckershospitalreview.com/hospital-management-administration/50-things-to-know-about-the-hospital-industry-2017.html
5http://www.beckershospitalreview.com/hospital-management-administration/50-things-to-know-about-the-hospital-industry-2017.html
6http://www.beckershospitalreview.com/quality/769-hospitals-see-medicare-payments-cut-over-high-hac-rates-7-things-to-know.html
7http://www.infectioncontroltoday.com/news/2017/01/sepsis-trumps-four-medical-conditions-tracked-by-cms-for-hospital-readmission-rates.aspx
