Data Analytics Addresses SDOH, Improves Length of Stay Rates

August 12, 2021 - By using data analytics, Parkview Medical Center and Pieces Technology addressed social determinants of health and decreased their average length of stay rates among patients.

Parkview Medical Center in Pueblo, Colorado, shared their patient load with one other hospital in the area. However, when the other hospital closed more than half its units, Parkview had to find ways to accommodate new and high volumes of patients.

However, Parkview Medical Center was experiencing high length of stay rates among patients. With high rates, Parkview could potentially not have the ability to accommodate for the new volume of patients, resulting in having to turn away individuals or transfer them to further hospitals.

“As you start digging into the data, you ask yourself what keeps patients in hospital? What causes a readmission to the hospital for a similar medical condition? Doctors and hospitals have the same antibiotics, we have the same standards of practice, we all do the same job, but what are the differences between us?” Parkview’s Chief Medical Officer Sandeep Vijan, MD, told HealthITAnalytics.

To ensure the hospital could provide care for the increased number of patients, Parkview examined the social determinants of health through data analytics to determine factors impacting population health and a patient’s length of stay.  

READ MORE: Improving Social Determinants of Health with Predictive Analytics

“It’s whether you have adequate insurance. It is whether you have social support at home. Whether you have food insecurity, transportation problems, unemployment, homelessness. These are the things in the fabric of our society today that really determine healthcare outcomes,” Vijan said.

Research indicated that social determinants of health have shown to be especially prevalent in Pueblo County.

According to a Common Wealth Fund report, many of the region’s residents have lived in general poverty for over a decade, with 46 percent living on incomes less than twice the federal poverty level. Additionally, one in four adults smoke, and almost a third are obese.

Smoking and obesity are both considered risk factors for diabetes, twice as preventable in adults in Pueblo County.

“Many residents suffer from disabilities, whether from the complications of disease, injuries from steel jobs, or self-harm due to substance abuse. Community leaders say that while Colorado’s legalization of marijuana has added jobs, it has also strained social services by attracting homeless and drug-seeking populations to the region, where the cost of living is lower than in other parts of the state,” the report stated.

READ MORE: Big Data Analytics, Social Determinants Reveal Heart Health Risks

In addition to addressing the social determinants of health, Parkview also evaluated data regarding the avenge length of stay time in their hospital. To meet the needs of new patients coming to the hospital, Parkview began exploring ways to decrease the length of stay times.

Length of stay is managed by following the patient through the entire care process, beginning from admission to discharge. To make the inpatient care process more efficient, providers establish an estimated date of discharge as early as possible. They can then streamline the discharge plan process, reducing the average length of stay and providing treatment to more patients.

To address issues regarding social determinants of health and length of stay, Parkview Medical Center developed a partnership with Pieces Technology to expand their artificial intelligence capabilities and provide more efficient care for the hospital’s patients.

“Pieces Technology integrates directly with our MEDITECH electronic medical records and natural language processing to read physician and nursing notes.  What that technology does is it allows the unseen elements of the patient’s care to be seen by the healthcare team,” Vijan explained.

The Pieces Technology system uses natural language processing, predictive modeling, machine learning, and AI to assist in the clinical decision-making process regarding patient care to reduce the length of stay rates.

READ MORE: Addressing the Social Determinants of Health with AI, Partnerships

The system’s algorithm examines unstructured data from electronic health records to determine key discharge barriers throughout the patient’s stay and creates a discharge checklist for physicians.

“In addition to harvesting the social determinants of health data, the AI platform has a proprietary algorithm that reads our progress notes from our physicians and is able to predict this person’s going to need oxygen at discharge, or this patient is high risk for readmission. It’s able to identify those factors and prevent them or present them to our healthcare team of physicians and nurses who participate in that care,” Vijan said.

Through Parkview Medical Center’s partnership with Pieces Technology, the hospital has significantly decreased their average length of stay as well as their excess length of stay, per national benchmark averages.

“We had high rates before, and we reduced that by 88 percent. Now we’re in the negative and better than the national average. What that means for us in the real world is our length of stay went from over five days to 4.45 days on average. That’s an absolute 6 percent increase in our efficiency for the year, which just means we can accept more patients,” Vijan said.