Predicting dental caries outcomes in young adults using machine learning approach.

OBJECTIVES: To predict the dental caries outcomes in young adults from a set of longitudinally-obtained predictor variables and identify the most important predictors using machine learning techniques. METHODS: This study was conducted using the Iowa Fluoride Study dataset. The predictor variables - sex, mother's education, family income, composite socio-economic status (SES), caries experience at ages 9, 13, and 17, and the cumulative estimates of risk and protective factors, including fluoride, dietary, and behavioral variables from ages 5-9, 9-13, 13-17, and 17-23 were used to predict the age 23 D2+MFS count. The following machine learning models (LASSO regression, generalized boosting machines (GBM), negative binomial (NegGLM), and extreme gradient boosting models (XGBOOST)) were compared under 5-fold cross validation with nested resampling techniques. RESULTS: The prevalence of cavitated level caries experience at age 23 (mean D2+MFS count) was 4.75. The predictive analysis found LASSO to be the best performing model (compared to GBM, NegGLM, and XGBOOST), with a root mean square error (RMSE) of 0.70, and coefficient of determination (R2) of 0.44. After dichotomization of the predicted and observed values of the LASSO regression, the classification results showed accuracy, precision, recall, and ROC AUC of 83.7%, 85.9%, 93.1%, 68.2%, respectively. Previous caries experience at age 13 and age 17 and sugar-sweetened beverages intakes at age 13 and age 17 were found to be the four most important predictors of cavitated caries count at age 23. CONCLUSION: Our machine learning model showed high accuracy and precision in the prediction of caries in young adults from a longitudinally-obtained predictor variables. Our model could, in the future, after further development and validation with other diverse population data, be used by public health specialists and policy-makers as a screening tool to identify the risk of caries in young adults and apply more targeted interventions. However, data from a more diverse population are needed to improve the quality and generalizability of caries prediction.

Causes of Death in the Presence of Law Enforcement in Johnson County, Iowa, 2011-2020.

Objectives. To examine sudden and unexpected or trauma-related deaths that occurred in the presence of law enforcement in Johnson County, Iowa, between 2011 and 2020. Methods. We identified deaths in the presence of law enforcement using definitions from the National Association of Medical Examiners. We obtained data, including demographics, cause and manner of death, toxicology results, and circumstances and location of event leading to death, from comprehensive medical examiner investigative reports. Results. There were 165 deaths that occurred in the presence of law enforcement: 114 were from a known disease, and 51 were either trauma related or the sudden, unexpected initial presentation of a previously unrecognized disease. Three deaths occurred in the context of physical restraint by law enforcement. Suicide was the leading manner of death among trauma-related deaths; the means of suicide was predictable based on in-custody (hanging) or precustody (firearm) circumstances. Conclusions. Our findings highlight the potential role of medical examiners and coroners in improving completeness of data on reporting death in the presence of law enforcement to public health agencies. (Am J Public Health. 2024;114(6):642-650. https://doi.org/10.2105/AJPH.2024.307616).

Summary of Roadway Incidents Involving Farm Equipment in Five Midwestern States Using the Fatality Analysis Reporting System (FARS).

Roadway incidents involving farm equipment is a growing area of concern among agricultural safety and health and public health professionals. The aim of this project was to evaluate the usefulness of the Fatality Analysis Reporting System (FARS) and analyze the number of roadway fatal incidents that involve farm equipment. Data collected from the FARS through the National Highway Traffic Safety Administration was used to summarize roadway incidents involving farm equipment. Cases from five midwestern states were analyzed from January to December 2020 using SPSS. Incidents involving farm equipment resulted in 25 cases with Iowa, Michigan, and Wisconsin all reporting six cases each. The most common manner of incidents were single-vehicle crashes and rear-ending incidents. Most of the events occurred during busy agricultural seasons, most often occurring in June and August with five cases each. The FARS dataset is a useful tool to identify cases, but it faces limitations, such as only reporting fatalities and lack of information on specific farm equipment involved in incidents. The results from the study are helpful to better understand roadway incidents and guide future intervention strategies.

Pesticide use and inflammatory bowel disease in licensed pesticide applicators and spouses in the Agricultural Health Study.

BACKGROUND: Pesticide exposure has been linked to some autoimmune diseases and colorectal cancer, possibly via alteration of gut microbiota or other mechanisms. While pesticides have been linked to gut dysbiosis and inflammation in animal models, few epidemiologic studies have examined pesticides in relation to inflammatory bowel disease (IBD). OBJECTIVES: We evaluated use of pesticides and incident IBD in 68,480 eligible pesticide applicators and spouses enrolled in the Agricultural Health Study. METHODS: Self-reported IBD cases were identified from follow-up questionnaires between enrollment (1993-1997) and 2022. We evaluated IBD incidence in relation to self-reported ever use of 50 pesticides among applicators and spouses. We also explored associations with intensity-weighted lifetime days (IWLD) of pesticide use among male applicators. Covariate-adjusted hazard ratios (HR) and 95% confidence intervals (CI) were calculated using Cox regression. RESULTS: We identified 454 IBD cases, including 227 among male applicators. In analyses with applicators and spouses combined, associations were positive (HR > 1.2) for ever vs. never use of five organochlorine insecticides, three organophosphate insecticides, one fungicide, and five herbicides. HRs were highest for dieldrin (HR = 1.59, 95%CI: 1.03, 2.44), toxaphene (HR = 1.61, 95%CI: 1.17, 2.21), parathion (HR = 1.42, 95%CI: 1.03, 1.95), and terbufos (HR = 1.53, 95%CI: 1.19, 1.96). We had limited power in many IWLD of pesticide use analyses and did not find clear evidence of exposure-response trends; however, we observed elevated HRs in all tertiles of IWLD use of terbufos compared to never use (T1 vs. never use HR = 1.52, 95%CI: 1.03, 2.24; T2 vs. never use HR = 1.53, 95%CI: 1.04, 2.26; T3 vs. never use HR = 1.51, 95%CI: 1.03, 2.23). CONCLUSIONS: Exposure to specific pesticides was associated with elevated hazards of IBD. These findings may have public health importance given the widespread use of pesticides and the limited number of known modifiable environmental risk factors for IBD.

Effects of rurality and distance to care on perinatal outcomes over a 1-year period during the COVID-19 pandemic.

PURPOSE: Our aim was to investigate the roles of rurality and distance to care on adverse perinatal outcomes and COVID-19 seroprevalence at the time of delivery over a 1-year period. METHODS: Data were collected from the electronic medical record on all pregnant patients who delivered at a single, large, Midwest academic medical center over 1 year. Rurality was classified using standard Rural-Urban Commuting Area codes. Geographic Information System tools were used to map outcomes. Data were analyzed with univariate and multivariate models, controlling for Body Mass Index (BMI), insurance status, and parity. FINDINGS: A total of 2,497 patients delivered during the study period; 20% of patients were rural (n = 499), 18.6% were micropolitan (n = 466), and 61.4% were metropolitan (n = 1,532). 10.4% of patients (n = 259) were COVID-19 seropositive. Rural patients did not experience higher rates of any measured adverse outcomes than metropolitan patients; micropolitan patients had increased odds of preterm labor (OR = 1.41, P = .022) and pre-eclampsia (OR = 1.78, P<.001). Patients living 30+ miles away from the medical center had increased odds of preterm labor (OR = 1.94, P<.001), pre-eclampsia (OR = 1.73, P = .002), and infant admission to the neonatal intensive care unit (OR = 2.12, P<.001), as well as lower gestational age at delivery (ß = -9.2 days, P<.001) and birth weight (ß = -206 grams, P<.001). CONCLUSION: Distance to care, rather than rurality, was the key predictor of multiple adverse perinatal outcomes in this cohort of deliveries over a 1-year period. Our study suggests that rurality should not be used as a standalone indicator of access to care without further knowledge of the specific barriers affecting a given population.