KANSAS CITY, MO. – The use of a socioeconomic-based risk index early in the hospital encounter can identify children at significantly increased risk of asthma readmission and family financial strain, a secondary analysis suggests.
The study of 601 patients with asthma found that children classified at high social risk had more than double the risk of hospital readmission or return emergency department visit within 12 months than did those at low risk, according to Dr. Andrew Beck, a pediatrics fellow at Cincinnati Children’s Hospital Medical Center, and his colleagues.
In addition, patients at high risk had nearly 15-fold increased odds of reporting two or more financial hardships.
Patrice Wendling/Elsevier Global Medical News
Dr. Andrew Beck
Dr. Beck observed that area-based geographic data is used routinely in public health for surveillance, resource allocation, and deprivation assessment but not at the bedside, where electronic medical records can improve data linkages from the moment patients register with their address.
"We don’t routinely use this data to identify patients who may be at increased social risk," he said during the plenary session at Pediatric Hospital Medicine 2011. "Every child receives the same basic, acute-oriented medical care, potentially missing the opportunity to inform other hospital, social, and medical interventions."
To illustrate his point, Dr. Beck pointed to a 20-fold difference in asthma admissions among patients aged 1-17 years and a 10-fold difference in median household income between Cincinnati’s tony, east-side Hyde Park neighborhood and the historic, urban Over-the-Rhine neighborhood. Some Cincinnati neighborhoods have no residents living in poverty, while others have 50% of children living below the poverty line.
The researchers devised a point-of-first-contact (PFC) risk index using the patient’s insurance status and U.S. census tract variables of poverty rate, home value, and education that were geographically coded to the patient’s address and zip code. One point was given if patients were on the "at-risk" side of the national median for each variable and another given if they were publicly insured or uninsured. Race/ethnicity was not included because of the potential for misclassification.
Analyzing the Backgrounds of Asthmatic Children
Based on the index, 117 of the 601 children, aged 1-16 years, in the prospective cohort, were at low risk (0 points), 201 were at medium risk (1-2 points), and 283 at high risk (3-4 points), he said at the meeting, sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and Academic Pediatric Association.
A total of 237 children (39%) returned to the ED or were readmitted within 12 months. Overall, 73% of the low-risk group had not returned to the hospital, compared with 64% of the medium-risk and 53% of the high-risk group, he said.
When this association was assessed in a Cox model using the low-risk group as the reference, the risk of reutilization was increased 70% among those at medium risk (hazard ratio, 1.7) and more than double among those at high risk (HR, 2.3).
The researchers then measured family financial strain in face-to-face interviews with caregivers using five validated questions chosen for their likelihood of leading to intervention. In all, 38% of families reported not having enough money to make ends meet, 11% said they did not have enough food to eat, 23% were unable to pay the full rent/mortgage, 39% were unable to pay full utilities, and 13% were forced to move in with others for financial reasons.
Nearly one-third (32%) of the sample answered "yes" to two or more financial strain questions. The odds of reporting two or more financial hardships were 14.8-fold higher for the high-risk group and 7.9-fold higher for the medium-risk group, which was statistically significant (both P value less than .0001), Dr. Beck said. Interestingly, this correlation had a high sensitivity (97%) and negative predictive value (95%), illustrating the index’s value as a screening tool, he added.
"Because the PFC index is based on data available so early on in the admission encounter and triage, we believe it could be used to quickly triage who may and who may not benefit from further assessment and intervention by deriving a risk profile before you even enter the room," Dr. Beck concluded.
The Value in Determining Patient Risk
Session co-moderator Dr. Karen Wilson of the University of Rochester (N.Y.) agreed that the data could be readily available but said more work needs to be done to determine how patients would react if told by an ED physician they were at risk based on where they live.
"It is a novel idea that I think deserves further investigation, but it’s probably not something, just based on our IT system, that is practical to use right now because we don’t have the computer systems to automatically generate that census tract information," she said in an interview.