• Digital Humanities and Cultural Heritage
• Publications close
• Canadian Institutes of Health Resea... close

The consumption and availability of ultra-processed foods (UPFs), which are associated with an increased risk of noncommunicable diseases, have increased in most countries. While many countries have or are planning to incorporate UPF recommendations in their national dietary guidelines, the classification of food processing levels relies on expertise-based manual categorization, which is labor-intensive and time-consuming. Our study utilized transformer-based language models to automate the classification of food processing levels according to the NOVA classification system in the Canada, Argentina, and US national food databases. We showed that fine-tuned language models using the ingredient list text found on food labels as inputs achieved a high overall accuracy (F1 score of 0.979) in predicting the food processing levels of Canadian food products, outperforming traditional machine learning models using structured nutrient data and bag-of-words. Most of the food categories reached a prediction accuracy of 0.98 using a fined-tuned language model, especially for predicting processed foods and ultra-processed foods. Our automation strategy was also effective and generalizable for classifying food products in the Argentina and US databases, providing a cost-effective approach for policymakers to monitor and regulate the UPFs in the global food supply.

Abstract Objectives Do adults without kin experience a care gap where they need help with activities of daily living but get no help from any source? We examine the prevalence of the care gap across Europe, and test whether those without partners or children substitute for their lack of close kin with help from broader networks, or whether they disproportionately experience care gaps. Methods Using data from the Survey on Health, Ageing and Retirement in Europe, we estimate the care gap in 28 European countries and Israel, how it varies, and who provides help for respondents with different family structures. Results The care gap is substantial, with 6.1% of all respondents ages 50 and above reporting a gap. It is highest in Western and Eastern Europe and lowest in Southern Europe and Israel. Respondents without partners or children are significantly more likely to have care gaps than those with close kin. However, respondents without close kin draw more often on more diverse networks of friends and relatives and use nursing home care. Discussion Our study introduces the concept of the care gap and shows that although it is most common among unpartnered adults without children it is also quite common for those with immediate family. A broader network partially but not completely substitutes for care gaps among those without immediate family. Our results offer new insights into the demand for public care services in countries with diverse welfare states.

AbstractThe ability to extract critical information about an infectious disease in a timely manner is critical for population health research. The lack of procedures for mining large amounts of health data is a major impediment. The goal of this research is to use natural language processing (NLP) to extract key information (clinical factors, social determinants of health) from free text. The proposed framework describes database construction, NLP modules for locating clinical and non-clinical (social determinants) information, and a detailed evaluation protocol for evaluating results and demonstrating the effectiveness of the proposed framework. The use of COVID-19 case reports is demonstrated for data construction and pandemic surveillance. The proposed approach outperforms benchmark methods in F1-score by about 1–3%. A thorough examination reveals the disease’s presence as well as the frequency of symptoms in patients. The findings suggest that prior knowledge gained through transfer learning can be useful when researching infectious diseases with similar presentations in order to accurately predict patient outcomes.

Indigenous Peoples are the fastest growing and youngest populations in Canada, yet face persistent housing crises across both urban and reserve settings. While Indigenous Peoples often move to cities in search of employment and education, they are overrepresented among the homeless population. This article, summarizing results from the author’s doctoral dissertation, provides a review of housing transitions needs and barriers from the experiences and perspectives of 13 Indigenous Elders/traditional knowledge keepers and youth, employing a narrative inquiry methodology and traditional knowledges interpretive framework. The intersection of Indigenous culture, community, housing transition, supports, and barriers is presented, highlighting the narrative voices, experiences, and perspectives of community members and Elders. Implications for service providers, policy, and future directions are included in this work.

Background: Despite the growth in research on intimate partner violence (IPV) as a chronic stressor, including studies on the negative mental health impacts of IPV, limited attention has been given to understanding the stress process in terms of the direct impact of IPV on mental health and the simultaneous mediating effects of social support (emotional or practical assistance from one’s network) and mastery (a sense of personal control) on the relationship between IPV and mental health [Depression symptoms and Post-traumatic stress disorder (PTSD) symptoms]. These mechanisms are also poorly understood in the context of the concurrent effects of age, mothering, education, and intimate partner status. Purposes: This dissertation examined the direct impact of IPV on mental health (Depression symptoms and PTSD symptoms) and the mechanisms that explain the relationship between severity of IPV and women’s mental health at a specific point in time, with a particular focus on testing whether and how social support and mastery mediated this relationship, in the context of selected conditions of women’s lives. The specific purposes were: (i) to examine whether mastery and social support mediate the effects of IPV severity on mental health and (ii) to identify which aspects of the social context of women’s lives account for variations in the mechanisms between IPV and mental health. Methods: Drawing on feminist intersectionality, a secondary analysis was conducted using data from a sample of 462 women with histories of recent IPV (i.e., past 6 months) and who participated in the iCAN plan 4 safety trial. Participants completed online surveys comprised of standard self-report measures for each study variable: IPV severity (Composite Abuse Scale), Depression (Center for Epidemiologic Studies Depression Scale), PTSD (PTSD checklist, Civilian Version), Mastery (Pearlin’s Mastery Scale) and Social Support (Medical Outcomes Study Social Support Survey). Structural equation modeling techniques were used to examine the mediating effects of both social support and mastery on the relationships between IPV severity on both depression and PTSD symptoms, while examining the effects of selected covariates capturing conditions of women’s lives (e.g., education, age, whether living with partner and mothering a minor child) on the direct and indirect effects of IPV on mental health. Results: The four key results are: (1) there was overall support for both direct and indirect effects in the model although direct effects were stronger; these results suggest strong relationship between severity of IPV and mental health problems even in the presence of mediators; (2) mastery but not social support mediated the relationship between IPV and mental health (for both outcomes: depression and PTSD symptoms), supporting nonconcurrent mediating effects of social support and mastery on the relationship between IPV and mental health, as opposed to the simultaneous mediating effects of these resources proposed in the Stress Process Model; (3) mothering reduced the negative effects of IPV on mental health, and although this effect was weak, it was stronger for depression than for PTSD; thus, mothering may offer some protection against the negative effect of IPV on depression and PTSD symptoms; (4) based on a subgroup analyses conducted with women who were and were not mothering dependent children, different mechanisms may explain the impact of IPV on mental health for women based on mothering status: social support mediated the relationship between IPV and mental health in women who identified as mothers of dependent children, while mastery mediated this relationship in women who did not identify as mothers of dependent children. Conclusions: The dissertation study produced new evidence that mastery mediates the relationship between severity of IPV and women’s mental health. If mastery is eroded by IPV, efforts to help women regain control could also help strengthen women’s health. Further, mothering is an important context for understanding the varied impacts of IPV on mental health. This reinforces the importance of taking an intersectional perspective and attempting to unpack the complexity of the mechanisms that explain the pathways between IPV, resources and mental health in ways that account for women’s differing realities.

Abstract Background Abstracting cerebrovascular disease (CeVD) from inpatient electronic medical records (EMRs) through natural language processing (NLP) is pivotal for automated disease surveillance and improving patient outcomes. Existing methods rely on coders’ abstraction, which has time delays and under-coding issues. This study sought to develop an NLP-based method to detect CeVD using EMR clinical notes. Methods CeVD status was confirmed through a chart review on randomly selected hospitalized patients who were 18 years or older and discharged from 3 hospitals in Calgary, Alberta, Canada, between January 1 and June 30, 2015. These patients’ chart data were linked to administrative discharge abstract database (DAD) and Sunrise™ Clinical Manager (SCM) EMR database records by Personal Health Number (a unique lifetime identifier) and admission date. We trained multiple natural language processing (NLP) predictive models by combining two clinical concept extraction methods and two supervised machine learning (ML) methods: random forest and XGBoost. Using chart review as the reference standard, we compared the model performances with those of the commonly applied International Classification of Diseases (ICD-10-CA) codes, on the metrics of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Result Of the study sample (n = 3036), the prevalence of CeVD was 11.8% (n = 360); the median patient age was 63; and females accounted for 50.3% (n = 1528) based on chart data. Among 49 extracted clinical documents from the EMR, four document types were identified as the most influential text sources for identifying CeVD disease (“nursing transfer report,” “discharge summary,” “nursing notes,” and “inpatient consultation.”). The best performing NLP model was XGBoost, combining the Unified Medical Language System concepts extracted by cTAKES (e.g., top-ranked concepts, “Cerebrovascular accident” and “Transient ischemic attack”), and the term frequency-inverse document frequency vectorizer. Compared with ICD codes, the model achieved higher validity overall, such as sensitivity (25.0% vs 70.0%), specificity (99.3% vs 99.1%), PPV (82.6 vs. 87.8%), and NPV (90.8% vs 97.1%). Conclusion The NLP algorithm developed in this study performed better than the ICD code algorithm in detecting CeVD. The NLP models could result in an automated EMR tool for identifying CeVD cases and be applied for future studies such as surveillance, and longitudinal studies.