Juhn, Young J.

Abstract Objective Artificial intelligence (AI) models may propagate harmful biases in performance and hence negatively affect the underserved. We aimed to assess the degree to which data quality of electronic health records (EHRs) affected by inequities related to low socioeconomic status (SES), results in differential performance of AI models across SES. Materials and Methods This study utilized existing machine learning models for predicting asthma exacerbation in children with asthma. We compared balanced error rate (BER) against different SES levels measured by HOUsing-based SocioEconomic Status measure (HOUSES) index. As a possible mechanism for differential performance, we also compared incompleteness of EHR information relevant to asthma care by SES. Results Asthmatic children with lower SES had larger BER than those with higher SES (eg, ratio = 1.35 for HOUSES Q1 vs Q2–Q4) and had a higher proportion of missing information relevant to asthma care (eg, 41% vs 24% for missing asthma severity and 12% vs 9.8% for undiagnosed asthma despite meeting asthma criteria). Discussion Our study suggests that lower SES is associated with worse predictive model performance. It also highlights the potential role of incomplete EHR data in this differential performance and suggests a way to mitigate this bias. Conclusion The HOUSES index allows AI researchers to assess bias in predictive model performance by SES. Although our case study was based on a small sample size and a single-site study, the study results highlight a potential strategy for identifying bias by using an innovative SES measure.

Background Inhaled corticosteroids (ICSs) are important in asthma management, but there are concerns regarding associated risk of pneumonia. While studies in asthmatic adults have shown inconsistent results, this risk in asthmatic children is unclear. Objective Our aim was to determine the association of ICS use with pneumonia risk in asthmatic children. Methods A nested case-control study was performed in the Mayo Clinic Birth Cohort. Asthmatic children (<18 years) with a physician diagnosis of asthma were identified from electronic medical records of children born at Mayo Clinic from 1997 to 2016 and followed until 31 December 2017. Pneumonia cases defined by Infectious Disease Society of America were 1:1 matched with controls without pneumonia by age, sex and asthma index date. Exposure was defined as ICS prescription at least 90 days prior to pneumonia. Associations of ICS use, type and dose (low, medium and high) with pneumonia risk were analysed using conditional logistic regression. Results Of the 2108 asthmatic children eligible for the study (70% mild intermittent and 30% persistent asthma), 312 children developed pneumonia during the study period. ICS use overall was not associated with risk of pneumonia (adjusted OR: 0.94, 95% CI: 0.62 to 1.41). Poorly controlled asthma was significantly associated with the risk of pneumonia (OR: 2.03, 95% CI: 1.35 to 3.05; p<0.001). No ICS type or dose was associated with risk of pneumonia. Conclusion ICS use in asthmatic children was not associated with risk of pneumonia but poorly controlled asthma was. Future asthma studies may need to include pneumonia as a potential outcome of asthma management.

Background: We recently developed an index of socioeconomic status (SES) termed HOUSES ( HOU sing-based index of S ocio E conomic S tatus) to overcome the absence of SES measures in common data sources. HOUSES has been associated with various childhood health outcomes to a similar extent as other measures of SES. Whether it is associated with outcomes in adults is unknown. To address this question, we assessed whether HOUSES was associated with mortality after MI. Methods: We conducted a population-based cohort study among all Olmsted County, MN residents with MI who were prospectively identified using predetermined criteria from November 1, 2002 to May 31, 2006. HOUSES was formulated based on a z-score for housing value, actual square footage, and numbers of bedrooms and bathrooms and categorized into quartiles. The 1-year mortality rate post MI was compared among subjects with different SES by HOUSES based on a log-rank test. Results: Among 696 patients with MI; 623 (90%) were successfully geo-coded to address and real property data. Of these 623 subjects, 594 (95%) were white and 263 (42%) female; the mean age was 68±15 years. Spearman correlation coefficient between HOUSES and education level was 0.23 (p<0.001). Mortality rates at 1 year following MI were 20% (95%CI: 13 - 26), 18% (11 - 24), 14% (8 - 19), and 16% (10 - 21) for patients in the first (lowest SES), second, third, and fourth quartiles (highest SES) of the HOUSES index, respectively (p=0.032). This association between HOUSES and mortality after MI was primarily explained by age and comorbid conditions which were associated with HOUSES. Cardioprotective drugs such as aspirin, beta-blocker, and statins were not associated with HOUSES nor accounted for the association between HOUSES and mortality after MI. Conclusions: Despite the lack of a strong correlation with other SES measure (i.e. no redundancy), HOUSES may be useful for overcoming the paucity of conventional SES measures in commonly used datasets in epidemiological research concerning cardiovascular disease among adults.

Background: Little is known about the impact of socioeconomic status (SES) as a key element of social determinants of health on intensive care unit (ICU) outcomes for adults. Objective: We assessed whether a validated individual SES index termed HOUSES (HOUsing-based SocioEconomic status index) derived from housing features was associated with short-term outcomes of critical illness including ICU mortality, ICU-free days, hospital-free days, and ICU readmission. Methods: We performed a population-based cohort study of adult patients living in Olmsted County, Minnesota, admitted to 7 intensive care units at Mayo Clinic from 2011 to 2014. We compared outcomes between the lowest SES group (HOUSES quartile 1 [Q1]) and the higher SES group (HOUSES Q2-4). We stratified the cohort based on age (<50 years old and ≥50 years old). Results: Among 4134 eligible patients, 3378 (82%) patients had SES successfully measured by the HOUSES index. Baseline characteristics, severity of illness, and reason for ICU admission were similar among the different SES groups as measured by HOUSES except for larger number of intoxications and overdoses in younger patients from the lowest SES. In all adult patients, there were no overall differences in mortality, ICU-free days, hospital-free days, or ICU readmissions in patients with higher SES compared to lower SES. Among older patients (>50 years), those with higher SES (HOUSES Q2-4) compared to those with lower SES (HOUSES Q1) had lower mortality rates (hazard ratio = 0.72; 95% CI: 0.56-0.93; adjusted P = .01), increased ICU-free days (mean 1.08 days; 95% CI: 0.34-1.84; adjusted P = .004), and increased hospital-free days (mean 1.20 days; 95% CI: 0.45-1.96; adjusted P = .002). There were no differences in ICU readmission rates (OR = 0.74; 95% CI: 0.55-1.00; P = .051). Conclusion: Individual-level SES may be an important determinant or predictor of critical care outcomes in older adults. Housing-based socioeconomic status may be a useful tool for enhancing critical care research and practice.

Introduction The lack of effective, consistent, reproducible and efficient asthma ascertainment methods results in inconsistent asthma cohorts and study results for clinical trials or other studies. We aimed to assess whether application of expert artificial intelligence (AI)-based natural language processing (NLP) algorithms for two existing asthma criteria to electronic health records of a paediatric population systematically identifies childhood asthma and its subgroups with distinctive characteristics. Methods Using the 1997–2007 Olmsted County Birth Cohort, we applied validated NLP algorithms for Predetermined Asthma Criteria (NLP-PAC) as well as Asthma Predictive Index (NLP-API). We categorised subjects into four groups (both criteria positive (NLP-PAC+/NLP-API+); PAC positive only (NLP-PAC+ only); API positive only (NLP-API+ only); and both criteria negative (NLP-PAC−/NLP-API−)) and characterised them. Results were replicated in unsupervised cluster analysis for asthmatics and a random sample of 300 children using laboratory and pulmonary function tests (PFTs). Results Of the 8196 subjects (51% male, 80% white), we identified 1614 (20%), NLP-PAC+/NLP-API+; 954 (12%), NLP-PAC+ only; 105 (1%), NLP-API+ only; and 5523 (67%), NLP-PAC−/NLP-API−. Asthmatic children classified as NLP-PAC+/NLP-API+ showed earlier onset asthma, more Th2-high profile, poorer lung function, higher asthma exacerbation and higher risk of asthma-associated comorbidities compared with other groups. These results were consistent with those based on unsupervised cluster analysis and lab and PFT data of a random sample of study subjects. Conclusion Expert AI-based NLP algorithms for two asthma criteria systematically identify childhood asthma with distinctive characteristics. This approach may improve precision, reproducibility, consistency and efficiency of large-scale clinical studies for asthma and enable population management.