An Illinois hospital system noticed something odd in their data: patients on Chicago's West Side were visiting the hospital for asthma-related complications at a far higher rate than anywhere else in their system.
In fact, they found, a handful of neighborhoods had the highest asthma hospitalization rate in the state. This pattern held true across all age ranges. Yet they did not have a higher baseline rate of asthma - only hospitalizations resulting from it. On average, these hospitalizations were costing $11,000 per patient, and as the patients were generally on Medicaid or uninsured, the hospital was expending millions of dollars per year.
Clinicians consider asthma hospitalizations to be generally avoidable with proper management. Yet after providing patients with superb treatment, subsidizing their medication, and empowering them with knowledge, they watched the same patients return to the hospital only weeks later.
This hotspot was not notable for anything they could think of that typically causes or exacerbates asthma. It was full of parks, far from highways, and immediately next to a world-class medical district.
Metopio gave the health system access to what they weren't able to see in their data: the community context.
Using Metopio’s analytics suite, the health system investigated correlations between the asthma hospitalizations and community conditions to identify likely root causes. By combining data from government agencies at the federal, state, and local level, Metopio made it easy to hone in on factors that set the West Side apart from other areas of Chicago - and that correlated strongly with asthma, even after controlling for other factors.
The prime suspect quickly became the old housing stock in the target neighborhoods - particularly buildings that had not been well-maintained since they were built in the early 20th century. A dataset of building violations provided the clearest evidence for this. Explore the scatterplot below showing this correlation.
Hover over the trend line for details. A regression analysis within Metopio shows that building violations have a highly significant relationship with asthma admissions, even after controlling for demographics and socioeconomic status.
The asthma hospitalizations proved to be even more highly correlated with lead poisoning in children, which is often caused by a lack of investment in maintaining old housing stock. Together, these results pointed to home conditions exacerbated by poorly maintained housing, such as untreated mold, broken HVAC systems, or pests.
Metopio moved the hospital from a reactive understanding of asthma to a contextualized and predictive approach. Using Metopio’s data answer, the hospital developed a pilot for a new model of care directed at the condition of the patient’s home.
Twenty homes were included initially. A patient who had been to the hospital for asthma in the past year was screened, offered in-home assessments and subsequent minor home improvements to address asthma triggers such as mold remediation, HVAC repairs, furnace cleaning, carpet replacement and integrated pest management.
The cost to screen a patient, inspect the home and remediate asthma triggers is $3,700 on average. This represents a $7,300 positive variance if it can avoid just one patient hospitalization for asthma.
Not one of the 20 patients in this initial group have been readmitted for asthma since the home remediation. Their outpatient asthma visits immediately fell by 80%.
By using Metopio to understand the community context and identify a root cause, the health system significantly improved patient wellness, reduced unnecessary asthma-related utilization of the hospital and saved the system money.
This pilot is now being used as the basis to talk with insurers about payment for the program, potentially turning the large cost of treating repeated asthma admissions into a reimbursable program that keeps the patients healthier and saves the broader healthcare system money.
This is just one way Metopio’s data answers can create scalable answers to your human questions. Let's talk about how to use an approach like this in your health system.