In This Article
- Data analysis is a useful tool for improving the cost, quality and delivery of health care throughout the United States
- The index hospital readmission rate following percutaneous coronary intervention at Mass General dropped 40% in four years—from 9.6% to 5.3%—after interventions gleaned by using natural language processing algorithms
- Electronic consults and mobile observation units improve health care delivery while reducing costs
At close to 18% of the gross national product, the cost of health care in the United States is unsustainable over the long term. “With clinical outcomes that are at best mediocre, we think we can do better,” says Jason H. Wasfy, MD, medical director at Massachusetts General Physicians Organization and director of quality and analytics at the Massachusetts General Hospital Corrigan Minehan Heart Center.
Dr. Wasfy sees data as the key to unlocking information that can influence change. He and his team are using data analysis of cardiology care to improve the cost, quality and delivery of health care in Massachusetts and throughout the United States.
One such analysis led to a 40% reduction in hospital readmission rates following percutaneous coronary intervention (PCI). The reduction was the result of using risk prediction, data from natural languge processing searches of medical records and a data set of statewide hospital admissions to inform changes in care process.
According to Dr. Wasfy, that’s just a hint of the care improvements possible with careful data analysis.
Managing Hospital Readmission Rates After PCI
Through multiple research efforts, Dr. Wasfy has analyzed large and small data sets to gain a better view of potential areas for improvement in health care. As a cardiologist, he knows the high costs of his field, and his work has largely focused on improvements to cardiology care. One such improvement is managing hospital readmissions, a common culprit of skyrocketing health care costs.
“Reducing hospital readmissions is a very attractive target because it offers hospitals the chance to simultaneously improve the quality of care while also reducing unnecessary costs,” says Dr. Wasfy.
Using data techniques and analysis, Dr. Wasfy and his team explored the causes of hospital readmission after cardiovascular procedures and cardiology diagnoses like myocardial infarction (MI). In one study, Dr. Wasfy and his team analyzed data from the Massachusetts Department of Health about PCI admissions performed at nonfederal hospitals. They aimed to predict which patients are more likely to come back to the hospital after PCI.
Their analysis identified several factors that put patients at higher risk for readmission after PCI, including:
- History of heart failure
- Previous MI
- Certain medical conditions, including diabetes, cerebrovascular disease, peripheral vascular disease and chronic lung disease
Critically, Dr. Wasfy’s research showed that patients are more likely to be readmitted for recurrent chest pain than true PCI complications, often leading to unnecessary diagnostic imaging and high costs.
A better understanding of what leads to readmissions after PCI helped researchers develop prediction models and risk scores that drove targeted interventions, including:
- A discharge checklist that ensures patients can access the medications they need and flags patients who need closer follow-up
- Educational videos that help patients understand normal vs. abnormal symptoms after PCI
- New avenues for follow-up care, such as electronic consultations
Implementation of these interventions at Mass General reduced the index hospital readmission rate 40% in four years—from 9.6% to 5.3%.
While these interventions proved extremely successful for reducing hospital readmissions after PCI, Dr. Wasfy found that the readmission rate after acute MI was not falling.
After another set of data analyses, the research team determined that they were looking at two different types of patients. Many patients who experienced an acute MI were not having PCIs. Researchers changed their approach and expanded tactics to include patients with acute MI who did not receive PCI. Dr. Wasfy says this approach has decreased the acute MI readmission rate by more than one-third.
Using Data Analytics to Improve Healthcare Delivery
Historically, hospital administrators and health outcomes researchers worked in parallel to improve the system in different ways. Dr. Wasfy sees his efforts as a small piece of a larger, learning health care system where administrators and researchers share information about program implementation and quality improvement efforts.
“What we've tried to do is to use the same standards that we would typically use in research efforts combined with what we've learned from hospital administrative efforts,” Dr. Wasfy explains. “By sharing our findings in academic publications, we hope the medical community—and especially patients throughout the country—will benefit.”
Dr. Wasfy and Mass General are also tackling the cost and quality equation in other ways, addressing care coordination and care transitions with novel care improvements. Electronic consultation is a prime example: “Does every patient need to see their cardiologist in the office for follow-up visits?” he asks. “In-office visits can be barriers to convenience while raising costs.”
Mass General has done 2,000 cardiology e-consults since 2014, with exceptional results. After rolling out the option of electronic consults—first in primary care and then in cardiology—patient satisfaction levels soared.
“This is a way to make patients happier, provide quicker access to medical expertise from specialists and also reduce the cost of care. It’s one more example of an area where we can improve the quality of care, this time in the outpatient setting,” he says.
Deploying mobile observation units as an alternative to hospitalization is another example of a care improvement that betters health care delivery while reducing costs. A mobile observation unit evaluates or monitors a patient at home instead of in the hospital. While Dr. Wasfy admits that these units aren’t appropriate for patients who need a high level of care, he says they work well for those who need intensive coordination of care for a short time.
Long-Term Goals of Improving Health Care
Through his data analysis efforts, Dr. Wasfy hopes to ensure the highest quality care possible at Mass General, and he has a larger goal: generating new knowledge that leads to improved health care quality and value nationwide.
“I don’t think it’s an exaggeration to say that health care cost and quality are perhaps the largest long-term threat to the American economy—and cardiology in particular is a large portion of this," says Dr. Wasfy. "At Mass General, we view this as a national challenge. We’re working simultaneously to improve care for patients throughout America with heart disease while also addressing the challenge in the healthcare system and economy.”
By using data in new ways, Dr. Wasfy aims to contribute to a common solution that benefits physicians and patients across the country. “Using rigorous statistical techniques and data analysis in a way that generates new knowledge and informs our administrative efforts is more common in research than in running a hospital. We’re trying to mix these two worlds,” he says.
The team hopes their efforts lead to system-wide care improvements and possibly even a change in the culture of medicine—toward a continuous learning environment that’s marked by sharing data.
“If we continue to advance data science and shift the culture towards one that’s constantly learning, I’m confident patients would benefit in many ways,” Dr. Wasfy says.
Refer a patient to Mass General's Heart Center