Browsing by Subject "Social risk factors"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Developing a brief assessment of social risks for the Veterans Health Administration Survey of Healthcare Experiences of Patients
    (Wiley, 2023) Hausmann, Leslie R. M.; Cohen, Alicia J.; Eliacin, Johanne; Gurewich, Deborah A.; Lee, Richard E.; McCoy, Jennifer L.; Meterko, Mark; Michaels, Zachary; Moy, Ernest M.; Procario, Gregory T.; Russell, Lauren E.; Schaefer, James H., Jr.; Medicine, School of Medicine
    Objective: To determine whether a 6- or 12-month look-back period affected rates of reported social risks in a social risk survey for use in the Veterans Health Administration and to assess associations of social risks with overall health and mental health. Study design: Cross-sectional survey of respondents randomized to 6- or 12-month look-back period. Data sources and study setting: Online survey with a convenience sample of Veterans in June and July 2021. Data collection/extraction methods: Veteran volunteers were recruited by email to complete a survey assessing social risks, including financial strain, adult caregiving, childcare, food insecurity, housing, transportation, internet access, loneliness/isolation, stress, discrimination, and legal issues. Outcomes included self-reported overall health and mental health. Chi-squared tests compared the prevalence of reported social risks between 6- and 12-month look-back periods. Spearman correlations assessed associations among social risks. Bivariate and multivariable logistic regression models estimated associations between social risks and fair/poor overall and mental health. Principal findings: Of 3418 Veterans contacted, 1063 (31.10%) responded (87.11% male; 85.61% non-Hispanic White; median age = 70, interquartile range [IQR] = 61-74). Prevalence of most reported social risks did not significantly differ by look-back period. Most social risks were weakly intercorrelated (median |r| = 0.24, IQR = 0.16-0.31). Except for legal issues, all social risks were associated with higher odds of fair/poor overall health and mental health in bivariate models. In models containing all significant social risks from bivariate models, adult caregiving and stress remained significant predictors of overall health; food insecurity, housing, loneliness/isolation, and stress remained significant for mental health. Conclusions: Six- and 12-month look-back periods yielded similar rates of reported social risks. Although most individual social risks are associated with fair/poor overall and mental health, when examined together, only adult caregiving, stress, loneliness/isolation, food, and housing remain significant.
  • Thumbnail Image
    Item
    Generalizability and portability of natural language processing system to extract individual social risk factors
    (Elsevier, 2023) Magoc, Tanja; Allen, Katie S.; McDonnell, Cara; Russo, Jean-Paul; Cummins, Jonathan; Vest, Joshua R.; Harle, Christopher A.; Emergency Medicine, School of Medicine
    Objective: The objective of this study is to validate and report on portability and generalizability of a Natural Language Processing (NLP) method to extract individual social factors from clinical notes, which was originally developed at a different institution. Materials and methods: A rule-based deterministic state machine NLP model was developed to extract financial insecurity and housing instability using notes from one institution and was applied on all notes written during 6 months at another institution. 10% of positively-classified notes by NLP and the same number of negatively-classified notes were manually annotated. The NLP model was adjusted to accommodate notes at the new site. Accuracy, positive predictive value, sensitivity, and specificity were calculated. Results: More than 6 million notes were processed at the receiving site by the NLP model, which resulted in about 13,000 and 19,000 classified as positive for financial insecurity and housing instability, respectively. The NLP model showed excellent performance on the validation dataset with all measures over 0.87 for both social factors. Discussion: Our study illustrated the need to accommodate institution-specific note-writing templates as well as clinical terminology of emergent diseases when applying NLP model for social factors. A state machine is relatively simple to port effectively across institutions. Our study. showed superior performance to similar generalizability studies for extracting social factors. Conclusion: Rule-based NLP model to extract social factors from clinical notes showed strong portability and generalizability across organizationally and geographically distinct institutions. With only relatively simple modifications, we obtained promising performance from an NLP-based model.