- Predictive analytics fueled by data from the Internet of Things (IoT) could help providers detect patients at high risk of emergency department (ED) visits and 30-day hospital readmissions, according to a study published in JMIR Medical Informatics.
The prevalence of chronic disease, along with the increasing elderly population, has led to a spike in costly ED visits and subsequent hospital admissions in the US. The research team noted that nearly 40 percent of patients visiting the ED arrive by emergency ambulance transport, and 80 percent of unscheduled hospital admissions occurred through the ED.
Organizations seeking to prevent avoidable ED visits and hospital admissions can leverage data from the devices that comprise the Internet of Things, such as personal emergency response systems (PERS). A PERS wearable device allows older patients to get help in a situation that could potentially require emergency transport by ambulance to the hospital, including sudden worsening of a chronic condition or a fall.
Patients can press the help button on their PERS device to connect to a call center associate, who can coordinate the necessary assistance.
When data from these devices is combined with user enrollment data, including information on patients’ medical conditions and demographics, it can provide valuable information about a patient’s status.
Researchers used retrospective PERS data to create a predictive analytics model that would flag individuals at high risk of emergency hospital transport within 30 days. The team found that that the predictive model achieved an area under the curve (AUC) of 0.779 when identifying patients at high risk of emergency hospital transport within 30 days.
Among the top one percent of predicted high-risk patients, 25.5 percent had one or more emergency hospital transports in the next 30 days.
The model also produced probability scores for each patient, ranging from zero to 100 percent, to assess the risk of 30-day emergency hospital transport. Researchers found that the model-predicted probabilities matched with observed patient outcomes. At predicted probabilities of 80 percent to 100 percent, 80 percent of patients experienced emergency hospital transport within one month.
The team then compared the accuracy of the model’s predictions with actual clinical outcomes derived from EHR data. The results showed that patients flagged as high risk by the model experienced 3.9 times more emergency encounters in the year following the prediction date than low-risk patients.
These findings indicate that using predictive models and data from IoT devices can help organizations reduce emergency room utilization while reducing costs. Providers can use this information to distribute limited healthcare resources to patients who need them the most, which could improve outcomes.
“Such risk assessment models have great potential to inform treatment decisions and improve the quality of care delivered to patients,” the team wrote.
“As innovative connected health technologies and value-based care policy influence the evolution of geriatric models of care to deliver more individualized, multifaceted management strategies, home health programs for community-dwelling seniors may benefit from enhanced risk assessment of patients.”
The study’s limitations included the fact that the PERS population was primarily made up of older female patients, and that individuals privately pay for the service, which isn’t covered by health insurance. These factors may limit the generalizability of the study.
Despite these limitations, the team expects that leveraging predictive analytics and data from remote devices can help providers target high-risk patients and reduce unnecessary care costs.
In future studies, the group will work to incorporate their predictive model into clinical workflows, allowing clinicians to see which patients are at highest risk of 30-day emergency transport. These patients will receive a nurse assessment and tailored interventions, which may prevent avoidable ED visits and hospital readmissions.
“By using remotely collected PERS data to predict the risk of emergency healthcare utilization, our predictive model of 30-day emergency hospital transport presents a unique opportunity to efficiently allocate limited health care resources to patients who need them the most and thereby reduce costs associated with excessive use,” the researchers said.
“Healthcare providers could benefit from our validated predictive model by estimating the risk of 30-day emergency hospital transport for individual patients and target timely preventive interventions to high-risk patients.”