Potential for Machine Learning to Address Data Gaps in Human Toxicity and Ecotoxicity Characterization.

Machine Learning (ML) is increasingly applied to fill data gaps in assessments to quantify impacts associated with chemical emissions and chemicals in products. However, the systematic application of ML-based approaches to fill chemical data gaps is still limited, and their potential for addressing a wide range of chemicals is unknown. We prioritized chemical-related parameters for chemical toxicity characterization to inform ML model development based on two criteria: (1) each parameter's relevance to robustly characterize chemical toxicity described by the uncertainty in characterization results attributable to each parameter and (2) the potential for ML-based approaches to predict parameter values for a wide range of chemicals described by the availability of chemicals with measured parameter data. We prioritized 13 out of 38 parameters for developing ML-based approaches, while flagging another nine with critical data gaps. For all prioritized parameters, we performed a chemical space analysis to assess further the potential for ML-based approaches to predict data for diverse chemicals considering the structural diversity of available measured data, showing that ML-based approaches can potentially predict 8-46% of marketed chemicals based on 1-10% with available measured data. Our results can systematically inform future ML model development efforts to address data gaps in chemical toxicity characterization.

Electronic cigarettes and health outcomes: umbrella and systematic review of the global evidence.

OBJECTIVE: To review and synthesise the global evidence regarding the health effects of electronic cigarettes (e-cigarettes, vapes). STUDY DESIGN: Umbrella review (based on major independent reviews, including the 2018 United States National Academies of Sciences, Engineering, and Medicine [NASEM] report) and top-up systematic review of published, peer-reviewed studies in humans examining the relationship of e-cigarette use to health outcomes published since the NASEM report. DATA SOURCES: Umbrella review: eight major independent reviews published 2017-2021. Systematic review: PubMed, MEDLINE, Scopus, Web of Science, the Cochrane Library, and PsycINFO (articles published July 2017 - July 2020 and not included in NASEM review). DATA SYNTHESIS: Four hundred eligible publications were included in our synthesis: 112 from the NASEM review, 189 from our top-up review search, and 99 further publications cited by other reviews. There is conclusive evidence linking e-cigarette use with poisoning, immediate inhalation toxicity (including seizures), and e-cigarette or vaping product use-associated lung injury (EVALI; largely but not exclusively for e-liquids containing tetrahydrocannabinol and vitamin E acetate), as well as for malfunctioning devices causing injuries and burns. Environmental effects include waste, fires, and generation of indoor airborne particulate matter (substantial to conclusive evidence). There is substantial evidence that nicotine e-cigarettes can cause dependence or addiction in non-smokers, and strong evidence that young non-smokers who use e-cigarettes are more likely than non-users to initiate smoking and to become regular smokers. There is limited evidence that freebase nicotine e-cigarettes used with clinical support are efficacious aids for smoking cessation. Evidence regarding effects on other clinical outcomes, including cardiovascular disease, cancer, development, and mental and reproductive health, is insufficient or unavailable. CONCLUSION: E-cigarettes can be harmful to health, particularly for non-smokers and children, adolescents, and young adults. Their effects on many important health outcomes are uncertain. E-cigarettes may be beneficial for smokers who use them to completely and promptly quit smoking, but they are not currently approved smoking cessation aids. Better quality evidence is needed regarding the health impact of e-cigarette use, their safety and efficacy for smoking cessation, and effective regulation. REGISTRATION: Systematic review: PROSPERO, CRD42020200673 (prospective).

Workforce participation in relation to cancer diagnosis, type and stage: Australian population-based study of 163,556 middle-aged people.

PURPOSE: To quantify the relationship of cancer diagnosis to workforce participation in Australia, according to cancer type, clinical features and personal characteristics. METHODS: Questionnaire data (2006-2009) from participants aged 45-64 years (n=163,556) from the population-based 45 and Up Study (n=267,153) in New South Wales, Australia, were linked to cancer registrations to ascertain cancer diagnoses up to enrolment. Modified Poisson regression estimated age- and sex-adjusted prevalence ratios (PRs) for non-participation in the paid workforce-in participants with cancer (n=8,333) versus without (n=155,223), for 13 cancer types. RESULTS: Overall, 42% of cancer survivors and 29% of people without cancer were out of the workforce (PR=1.18; 95%CI=1.15-1.21). Workforce non-participation varied substantively by cancer type, being greatest for multiple myeloma (1.83; 1.53-2.18), oesophageal (1.70; 1.13-2.58) and lung cancer (1.68; 1.45-1.93) and moderate for colorectal (1.23; 1.15-1.33), breast (1.11; 1.06-1.16) and prostate cancer (1.06; 0.99-1.13). Long-term survivors, 5 or more years post-diagnosis, had 12% (7-16%) greater non-participation than people without cancer, and non-participation was greater with recent diagnosis, treatment or advanced stage. Physical disability contributed substantively to reduced workforce participation, regardless of cancer diagnosis. CONCLUSIONS: Cancer survivors aged 45-64 continue to participate in the workforce. However, participation is lower than in people without cancer, varying by cancer type, and is reduced particularly around the time of diagnosis and treatment and with advanced disease. IMPLICATIONS FOR CANCER SURVIVORS: While many cancer survivors continue with paid work, participation is reduced. Workforce retention support should be tailored to survivor preferences, cancer type and cancer journey stage.