- Unplanned hospital readmissions are one of the costliest services in healthcare, with organizations shelling out billions each year on these frequently avoidable episodes.
With the rise of value-based care initiatives, most notably CMS’s Hospital Readmissions Reduction Program (HRRP), organizations are looking for ways to use their many sources of data to cut down on expensive readmission rates.
From electronic health records to newer data assets like genomic data, healthcare entities now have more materials than ever to generate actionable clinical insights.
What are some of the ways hospitals can use big data analytics to reduce hospital readmissions, lower costs, and improve health outcomes?
Develop risk stratification and predictive analytics capabilities
Each year, the healthcare industry spends billions of dollars on preventable services, including hospital readmissions. Reducing readmission rates will require organizations to deliver preventive, forward-thinking care to those patients who need it most.
By examining patient data, providers can start to see which factors will impact future health outcomes, and begin to develop risk scores and predictive algorithms to create tailored care interventions.
While some organizations may feel they lack the necessary data to build these capabilities, research has shown that entities can form predictive models with less-than-perfect data.
For example, a team at the University of Washington Tacoma developed a predictive analytics algorithm to flag 30-day readmissions for heart failure patients. The tool used several common clinical and demographic metrics, but could also function without the inclusion of certain variables, making it useful for providers and patients who may not have all of the data on hand.
Healthcare organizations can also use data elements that are typically overlooked to develop risk scores.
Advocate Healthcare, a Chicago-based health system, was able to cut readmissions and save hospitals more than $4.8 million by implementing a patient nutrition care program.
“Value-based care means looking comprehensively at patient care to identify gaps and opportunities for improvement,” said Lee Sacks, MD, executive vice president and chief medical officer of Advocate Health Care.
“The study's findings demonstrate that modest changes in the way we care for patients, such as ensuring patients are nourished during their hospital stay, can have a big impact in reducing costs and improving health outcomes.”
Leverage patient engagement technology
To avoid returning to the hospital, patients must understand how to stay healthy after they leave. While it can be difficult for any patient to follow post-discharge instructions, adhering to care plans can be especially complicated for patients with multiple chronic conditions, particularly if they don’t communicate with their providers.
The Agency for Healthcare Research and Quality (AHRQ) reports that nearly 20 percent of Medicare chronic disease patients were readmitted to a hospital within 30 days because their condition worsened.
Of those readmitted patients, nearly half had no post-discharge contact with healthcare professionals.
To reduce readmission rates, providers can use data-driven patient engagement tools to help individuals follow their treatment plans.
“Patients need to be educated about their treatment plan, self-management, and how to detect warning signs of problems,” Jack Meyer, PhD, Managing Principal, Health Management Associates said in an interview with AHRQ.
“Healthcare providers can work with patients to monitor chronic conditions from their homes through the use of electronic devices that transmit patient data, such as body weight or blood glucose levels, directly to the physicians' offices.”
Additionally, organizations can use mobile apps and messages to help patients with their post-discharge treatment plans.
A 2017 study showed that real-time mHealth messages helped patients stay on track with their medication, generating an engagement rate of 86 percent throughout the study period.
Utilize clinical decision support tools
The many sources of data that providers must consider when deciding how to best treat a patient can be overwhelming. Medical histories, current medications, and lifestyle choices all factor into these choices, and it is critical that clinicians examine all this information to ensure patients don’t end up back in the hospital.
To augment decision-making at the point of care, organizations can implement clinical decision support (CDS) tools that leverage cutting-edge data analytics technology.
Advancements in artificial intelligence and natural language processing are expected to drive a new generation of clinical decision support, with the potential to analyze large datasets and help reduce unnecessary healthcare utilization.
Organizations can also combine CDS with data not included in electronic health records to lower readmission rates, such as genetic sequencing results.
In a 2017 study, researchers used a CDS tool and genetic testing to treat home health patients using multiple medications. The results showed that they were able to reduce hospital readmissions by 52 percent.
Enhance care coordination, communication
The seamless exchange of health information is essential to ensure patients don’t re-enter the hospital after they leave. Patients must understand their treatment plans as they move between care settings and back to their homes, which requires care coordination and communication among clinicians.
However, delivering coordinated, comprehensive care throughout the clinical process and beyond is challenging. Disparate health IT systems and a lack of collaboration can pose significant barriers to care coordination.
Statewide health information exchanges (HIEs) could offer a viable solution to this issue. Participating providers could receive information about patients from separate organizations and improve care management.
In Rhode Island, an HIE is delivering real-time admission, discharge, and transfer (ADT) alerts to providers across the state, and has decreased hospital readmissions by 19 percent, saving the state $13.3 million.
Increased collaboration among nurses and physicians has also proven to reduce readmissions. At Brigham and Women’s Hospital, an organization-wide effort to identify and manage high-risk patients reduced preventable rehospitalizations by nine percent over three years.
Nurses and physicians worked closely with each other to discuss the best care approach for every individual, with careful consideration of each patient’s health history and symptoms.
Reducing hospital readmission rates will require organizations to leverage their big data resources to unlock actionable insights and ultimately improve patient care. Delivering preventive, comprehensive care will result in fewer hospital stays, reduced costs, and better health outcomes.