RESUMEN

BACKGROUND: Although one objective of NHS 111 is to ease the strain on urgent and emergency care services, studies suggest the telephone triage service may be contributing to increased demand. Moreover, while parents and caregivers generally find NHS 111 satisfactory, concerns exist about its integration with the healthcare system and the appropriateness of advice. This study aimed to analyse the advice provided in NHS 111 calls, the duration between the call and ED attendance, and the outcomes of such attendances made by children and young people (C&YP). METHODS: A retrospective cohort study was carried out of C&YP (≤17) attending an ED in the Yorkshire and Humber region of the UK following contact with NHS 111 between 1 April 2016 and 31 March 2017. This linked-data study examined NHS 111 calls and ED outcomes. Lognormal mixture distributions were fit to compare the time taken to attend ED following calls. Logistic mixed effects regression models were used to identify predictors of low-acuity NHS 111-related ED attendances. RESULTS: Our study of 348 401 NHS 111 calls found they were primarily concerning children aged 0-4 years. Overall, 13.1% of calls were followed by an ED attendance, with a median arrival time of 51 minutes. Of the 34 664 calls advising ED attendance 41% complied, arriving with a median of 38 minutes-27% of which defined as low-acuity. Although most calls advising primary care were not followed by an ED attendance (93%), those seen in an ED generally attended later (median 102 minutes) with 23% defined as low-acuity. Younger age (<1) was a statistically significant predictor of low-acuity ED attendance following all call dispositions apart from home care. CONCLUSION: More tailored options for unscheduled healthcare may be needed for younger children. Both early low-acuity attendance and late high-acuity attendance following contact with NHS 111 could act as useful entry points for clinical audits of the telephone triage service.

RESUMEN

BACKGROUND: It remains unknown to what extent intrinsic atrial cardiomyopathy or left ventricular diastolic dysfunction drive atrial remodeling and functional failure in heart failure with preserved ejection fraction (HFpEF). Computational 3-dimensional (3D) models fitted to cardiovascular magnetic resonance allow state-of-the-art anatomic and functional assessment, and we hypothesized to identify a phenotype linked to HFpEF. METHODS: Patients with exertional dyspnea and diastolic dysfunction on echocardiography (E/e', >8) were prospectively recruited and classified as HFpEF or noncardiac dyspnea based on right heart catheterization. All patients underwent rest and exercise-stress right heart catheterization and cardiovascular magnetic resonance. Computational 3D anatomic left atrial (LA) models were generated based on short-axis cine sequences. A fully automated pipeline was developed to segment cardiovascular magnetic resonance images and build 3D statistical models of LA shape and find the 3D patterns discriminant between HFpEF and noncardiac dyspnea. In addition, atrial morphology and function were quantified by conventional volumetric analyses and deformation imaging. A clinical follow-up was conducted after 24 months for the evaluation of cardiovascular hospitalization. RESULTS: Beyond atrial size, the 3D LA models revealed roof dilation as the main feature found in masked HFpEF (diagnosed during exercise-stress only) preceding a pattern shift to overall atrial size in overt HFpEF (diagnosed at rest). Characteristics of the 3D model were integrated into the LA HFpEF shape score, a biomarker to characterize the gradual remodeling between noncardiac dyspnea and HFpEF. The LA HFpEF shape score was able to discriminate HFpEF (n=34) to noncardiac dyspnea (n=34; area under the curve, 0.81) and was associated with a risk for atrial fibrillation occurrence (hazard ratio, 1.02 [95% CI, 1.01-1.04]; P=0.003), as well as cardiovascular hospitalization (hazard ratio, 1.02 [95% CI, 1.00-1.04]; P=0.043). CONCLUSIONS: LA roof dilation is an early remodeling pattern in masked HFpEF advancing to overall LA enlargement in overt HFpEF. These distinct features predict the occurrence of atrial fibrillation and cardiovascular hospitalization. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03260621.

Asunto(s)

RESUMEN

Clinical prediction models often aim to predict rare, high-risk events, but building such models requires robust understanding of imbalance datasets and their unique study design considerations. This practical guide highlights foundational prediction model principles for surgeon-data scientists and readers who encounter clinical prediction models, from feature engineering and algorithm selection strategies to model evaluation and design techniques specific to imbalanced datasets. We walk through a clinical example using readable code to highlight important considerations and common pitfalls in developing machine learning-based prediction models. We hope this practical guide facilitates developing and critically appraising robust clinical prediction models for the surgical community.

RESUMEN

OBJECTIVES: The impact of the systematic screening of Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) in men having sex with men (MSM) on these pathogens' epidemiology remains unclear. We conducted a modelling study to analyse this impact in French MSM. METHODS: We modelled NG and CT transmission using a site-specific deterministic compartmental model. We calibrated NG and CT prevalence at baseline using results from MSM enrolled in the Dat'AIDS cohort. The baseline scenario was based on 1 million MSM, 40 000 of whom were tested every 90 days and 960 000 every 200 days. Incidence rate ratios (IRRs) at steady state were simulated for NG, CT, NG and/or CT infections, for different combinations of tested sites, testing frequency and numbers of frequently tested patients. RESULTS: The observed prevalence rate was 11.0%, 10.5% and 19.1% for NG, CT and NG and/or CT infections. The baseline incidence rate was estimated at 138.2 per year per 100 individuals (/100PY), 86.8/100PY and 225.0/100PY for NG, CT and NG and/or CT infections. Systematically testing anal, pharyngeal and urethral sites at the same time reduced incidence by 14%, 23% and 18% (IRR: 0.86, 0.77 and 0.82) for NG, CT and NG and/or CT infections. Reducing the screening interval to 60 days in frequently tested patients reduced incidence by 20%, 29% and 24% (IRR: 0.80, 0.71 and 0.76) for NG, CT and NG and/or CT infections. Increasing the number of frequently tested patients to 200 000 reduced incidence by 29%, 40% and 33% (IRR: 0.71, 0.60 and 0.67) for NG, CT and NG and/or CT infections. No realistic scenario could decrease pathogens' incidence by more than 50%. CONCLUSIONS: To curb the epidemic of NG and CT in MSM, it would not only be necessary to drastically increase screening, but also to add other combined interventions.

Asunto(s)

RESUMEN

Background: Tiered trauma team activation (TTA) allows systems to optimally allocate resources to an injured patient. Target undertriage and overtriage rates of <5% and <35% are difficult for centers to achieve, and performance variability exists. The objective of this study was to optimize and externally validate a previously developed hospital trauma triage prediction model to predict the need for emergent intervention in 6 hours (NEI-6), an indicator of need for a full TTA. Methods: The model was previously developed and internally validated using data from 31 US trauma centers. Data were collected prospectively at five sites using a mobile application which hosted the NEI-6 model. A weighted multiple logistic regression model was used to retrain and optimize the model using the original data set and a portion of data from one of the prospective sites. The remaining data from the five sites were designated for external validation. The area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC) were used to assess the validation cohort. Subanalyses were performed for age, race, and mechanism of injury. Results: 14 421 patients were included in the training data set and 2476 patients in the external validation data set across five sites. On validation, the model had an overall undertriage rate of 9.1% and overtriage rate of 53.7%, with an AUROC of 0.80 and an AUPRC of 0.63. Blunt injury had an undertriage rate of 8.8%, whereas penetrating injury had 31.2%. For those aged ≥65, the undertriage rate was 8.4%, and for Black or African American patients the undertriage rate was 7.7%. Conclusion: The optimized and externally validated NEI-6 model approaches the recommended undertriage and overtriage rates while significantly reducing variability of TTA across centers for blunt trauma patients. The model performs well for populations that traditionally have high rates of undertriage. Level of evidence: 2.

RESUMEN

Objective: The objective of this study is to estimate the prevalence of chronic Chagas disease (CCD) in Brazil: in the general population, in women, and in women of childbearing age. Methods: A meta-analysis of the literature was conducted to extract data on the prevalence of CCD in municipalities in Brazil in the 2010-2022 period: in the general population, in women, and in women of childbearing age. Municipal-level CCD indicators available in health information systems were selected. Statistical modeling of the data extracted from the meta-analysis (based on data obtained from information systems) was applied to linear, generalized linear, and additive models. Results: The five most appropriate models were selected from a total of 549 models tested to obtain a consensus model (adjusted R2 = 54%). The most important predictor was self-reported CCD in the primary health care information system. Zero prevalence was estimated in 1 792 (32%) of Brazil's 5 570 municipalities; in the remaining 3 778 municipalities, average prevalence of the disease was estimated at 3.25% (± 2.9%). The number of carriers of CCD was estimated for the Brazilian population (~3.7 million), for women (~2.1 million) and for women of childbearing age (~590 000). The disease reproduction rate was calculated at 1.0336. All estimates refer to the 2015-2016 period. Conclusions: The estimated prevalence of CCD, especially among women of childbearing age, highlights the challenge of vertical transmission in Brazilian municipalities. Mathematical projections suggest that these estimates should be included in the national program for the elimination of vertical transmission of Chagas disease.

Objetivo: El objetivo de este estudio fue estimar la prevalencia de la enfermedad de Chagas crónica en la población brasileña en general, en las mujeres y en las mujeres en edad fértil. Métodos: Se realizó un metanálisis de la bibliografía para extraer datos sobre la prevalencia de la enfermedad de Chagas crónica en la población brasileña en general, en las mujeres y en las mujeres en edad fértil, en los municipios de Brasil durante el período 2010-2022. Se seleccionaron los indicadores relacionados con esa enfermedad disponibles en los sistemas municipales de información de salud. La modelización estadística de los datos extraídos del metanálisis, en función de los obtenidos de los sistemas de información, se aplicó a modelos lineales, lineales generalizados y aditivos. Resultados: Se seleccionaron los cinco modelos más apropiados de un total de 549 modelos evaluados, para obtener un modelo de consenso (R2 ajustado = 54%). El factor predictor más importante fue el registro de la enfermedad de Chagas crónica autodeclarada en el sistema de información de atención primaria de salud. De los 5570 municipios brasileños, en 1792 (32%) la prevalencia estimada fue nula y en los 3778 restantes la prevalencia media fue del 3,25% (± 2,9%). El número estimado de pacientes con enfermedad de Chagas crónica en la población brasileña en general, en las mujeres y en las mujeres en edad fértil fue de ~3,7 millones, ~2,1 millones y ~590 000, respectivamente. La tasa calculada de reproducción de la enfermedad fue de 1,0336. Todas las estimaciones se refieren al período 2015-2016. Conclusiones: La prevalencia estimada de la enfermedad de Chagas crónica, especialmente en las mujeres en edad fértil, pone de manifiesto el desafío que representa la transmisión vertical en los municipios brasileños. Estas estimaciones están en línea con los patrones de las proyecciones matemáticas, y sugieren la necesidad de incorporarlas al Pacto Nacional para la Eliminación de la Transmisión Vertical de la Enfermedad de Chagas.

RESUMEN

BACKGROUND AND PURPOSE: Concerns over chest wall toxicity has led to debates on treating tumors adjacent to the chest wall with single-fraction stereotactic ablative radiotherapy (SABR). We performed a secondary analysis of patients treated on the prospective iSABR trial to determine the incidence and grade of chest wall pain and modeled dose-response to guide radiation planning and estimate risk. MATERIALS AND METHODS: This analysis included 99 tumors in 92 patients that were treated with 25 Gy in one fraction on the iSABR trial which individualized dose by tumor size and location. Toxicity events were prospectively collected and graded based on the CTCAE version 4. Dose-response modeling was performed using a logistic model with maximum likelihood method utilized for parameter fitting. RESULTS: There were 22 grade 1 or higher chest wall pain events, including five grade 2 events and zero grade 3 or higher events. The volume receiving at least 11 Gy (V11Gy) and the minimum dose to the hottest 2 cc (D2cc) were most highly correlated with toxicity. When dichotomized by an estimated incidence of ≥ 20 % toxicity, the D2cc > 17 Gy (36.6 % vs. 3.7 %, p < 0.01) and V11Gy > 28 cc (40.0 % vs. 8.1 %, p < 0.01) constraints were predictive of chest wall pain, including among a subset of patients with tumors abutting or adjacent to the chest wall. CONCLUSION: For small, peripheral tumors, single-fraction SABR is associated with modest rates of low-grade chest wall pain. Proximity to the chest wall may not contraindicate single fractionation when using highly conformal, image-guided techniques with sharp dose gradients.

Asunto(s)

RESUMEN

Objectives: In medical research, the study's design and statistical methods are pivotal, as they guide interpretation and conclusion. Selecting appropriate statistical models hinges on the distribution of the outcome measure. Count data, frequently used in medical research, often exhibit over-dispersion or zero inflation. Occasionally, count data are considered ordinal (with a maximum outcome value of 5), and this calls for the application of ordinal regression models. Various models exist for analyzing over-dispersed data such as negative binomial, generalized Poisson (GP), and ordinal regression model. This study aims to examine whether the GP model is a superior alternative to the ordinal logistic regression (OLR) model, specifically in the context of zero-inflated Poisson models using both simulated and real-time data. Methods: Simulated data were generated with varied estimates of regression coefficients, sample sizes, and various proportions of zeros. The GP and OLR models were compared using fit statistics. Additionally, comparisons were made using real-time datasets. Results: The simulated results consistently revealed lower bias and mean squared error values in the GP model compared to the OLR model. The same trend was observed in real-time datasets, with the GP model consistently demonstrating lower standard errors. Except when the sample size was 1000 and the proportions of zeros were 30% and 40%, the Bayesian information criterion consistently favored the GP model over the OLR model. Conclusions: This study establishes that the proposed GP model offers a more advantageous alternative to the OLR model. Moreover, the GP model facilitates easier modeling and interpretation when compared to the OLR model.

RESUMEN

OBJECTIVE: To estimate the prevalence of long COVID among Western Australian adults, a highly vaccinated population whose first major exposure to the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was during the 2022 Omicron wave, and to assess its impact on health service use and return to work or study. STUDY DESIGN: Follow-up survey (completed online or by telephone). SETTING, PARTICIPANTS: Adult Western Australians surveyed 90 days after positive SARS-CoV-2 test results (polymerase chain reaction or rapid antigen testing) during 16 July - 3 August 2022 who had consented to follow-up contact for research purposes. MAIN OUTCOME MEASURES: Proportion of respondents with long COVID (ie, reporting new or ongoing symptoms or health problems, 90 days after positive SARS-CoV-2 test result); proportion with long COVID who sought health care for long COVID-related symptoms two to three months after infection; proportion who reported not fully returning to previous work or study because of long COVID-related symptoms. RESULTS: Of the 70 876 adults with reported SARS-CoV-2 infections, 24 024 consented to contact (33.9%); after exclusions, 22 744 people were invited to complete the survey, of whom 11 697 (51.4%) provided complete responses. Our case definition for long COVID was satisfied by 2130 respondents (18.2%). The risk of long COVID was greater for women (v men: adjusted risk ratio [aRR], 1.5; 95% confidence interval [CI], 1.4-1.6) and for people aged 50-69 years (v 18-29 years: aRR, 1.6; 95% CI, 1.4-1.9) or with pre-existing health conditions (aRR, 1.5; 95% CI, 1.4-1.7), as well as for people who had received two or fewer COVID-19 vaccine doses (v four or more: aRR, 1.4; 95% CI, 1.2-1.8) or three doses (aRR, 1.3; 95% CI, 1.1-1.5). The symptoms most frequently reported by people with long COVID were fatigue (1504, 70.6%) and concentration difficulties (1267, 59.5%). In the month preceding the survey, 814 people had consulted general practitioners (38.2%) and 34 reported being hospitalised (1.6%) with long COVID. Of 1779 respondents with long COVID who had worked or studied before the infection, 318 reported reducing or discontinuing this activity (17.8%). CONCLUSION: Ninety days after infection with the Omicron SARS-CoV-2 variant, 18.2% of survey respondents reported symptoms consistent with long COVID, of whom 38.7% (7.1% of all survey respondents) sought health care for related health concerns two to three months after the acute infection.

Asunto(s)

RESUMEN

OBJECTIVE: To test the ability of high-performance machine learning (ML) models employing clinical, radiological, and radiomic variables to improve non-invasive prediction of the pathological status of prostate cancer (PCa) in a large, single-institution cohort. METHODS: Patients who underwent multiparametric MRI and prostatectomy in our institution in 2015-2018 were considered; a total of 949 patients were included. Gradient-boosted decision tree models were separately trained using clinical features alone and in combination with radiological reporting and/or prostate radiomic features to predict pathological T, pathological N, ISUP score, and their change from preclinical assessment. Model behavior was analyzed in terms of performance, feature importance, Shapley additive explanation (SHAP) values, and mean absolute error (MAE). The best model was compared against a naïve model mimicking clinical workflow. RESULTS: The model including all variables was the best performing (AUC values ranging from 0.73 to 0.96 for the six endpoints). Radiomic features brought a small yet measurable boost in performance, with the SHAP values indicating that their contribution can be critical to successful prediction of endpoints for individual patients. MAEs were lower for low-risk patients, suggesting that the models find them easier to classify. The best model outperformed (p ≤ 0.0001) clinical baseline, resulting in significantly fewer false negative predictions and overall was less prone to under-staging. CONCLUSIONS: Our results highlight the potential benefit of integrative ML models for pathological status prediction in PCa. Additional studies regarding clinical integration of such models can provide valuable information for personalizing therapy offering a tool to improve non-invasive prediction of pathological status. CLINICAL RELEVANCE STATEMENT: The best machine learning model was less prone to under-staging of the disease. The improved accuracy of our pathological prediction models could constitute an asset to the clinical workflow by providing clinicians with accurate pathological predictions prior to treatment. KEY POINTS: • Currently, the most common strategies for pre-surgical stratification of prostate cancer (PCa) patients have shown to have suboptimal performances. • The addition of radiological features to the clinical features gave a considerable boost in model performance. Our best model outperforms the naïve model, avoiding under-staging and resulting in a critical advantage in the clinic. •Machine learning models incorporating clinical, radiological, and radiomics features significantly improved accuracy of pathological prediction in prostate cancer, possibly constituting an asset to the clinical workflow.

Asunto(s)

RESUMEN

OBJECTIVES: To project how many minimal trauma fractures could be averted in Australia by expanding the number and changing the operational characteristics of fracture liaison services (FLS). STUDY DESIGN: System dynamics modelling. SETTING, PARTICIPANTS: People aged 50 years or more who present to hospitals with minimal trauma fractures, Australia, 2020-31. MAIN OUTCOME MEASURES: Numbers of all minimal trauma fractures and of hip fractures averted by increasing the FLS number (from 29 to 58 or 100), patient screening rate (from 30% to 60%), and capacity for accepting new patients (from 40 to 80 per service per month), and reducing the proportion of eligible patients who do not attend FLS (from 30% to 15%); cost per fracture averted. RESULTS: Our model projected a total of 2 441 320 minimal trauma fractures (258 680 hip fractures; 2 182 640 non-hip fractures) in people aged 50 years or older during 2020-31, including 1 211 646 second or later fractures. Increasing the FLS number to 100 averted a projected 5405 fractures (0.22%; $39 510 per fracture averted); doubling FLS capacity averted a projected 3674 fractures (0.15%; $35 835 per fracture averted). Our model projected that neither doubling the screening rate nor reducing by half the proportion of eligible patients who did not attend FLS alone would reduce the number of fractures. Increasing the FLS number to 100, the screening rate to 60%, and capacity to 80 new patients per service per month would together avert a projected 13 672 fractures (0.56%) at a cost of $42 828 per fracture averted. CONCLUSION: Our modelling indicates that increasing the number of hospital-based FLS and changing key operational characteristics would achieve only moderate reductions in the number of minimal trauma fractures among people aged 50 years or more, and the cost would be relatively high. Alternatives to specialist-led, hospital-based FLS should be explored.

Asunto(s)

RESUMEN

Traumatic brain injury is a leading cause of death and disability worldwide. Interventions that mitigate secondary brain injury have the potential to improve outcomes for patients and reduce the impact on communities and society. Increased circulating catecholamines are associated with worse outcomes and there are supportive animal data and indications in human studies of benefit from beta-blockade after severe traumatic brain injury. Here, we present the protocol for a dose-finding study using esmolol in adults commenced within 24 h of severe traumatic brain injury. Esmolol has practical advantages and theoretical benefits as a neuroprotective agent in this setting, but these must be balanced against the known risk of secondary injury from hypotension. The aim of this study is to determine a dose schedule for esmolol, using the continual reassessment method, that combines a clinically significant reduction in heart rate as a surrogate for catecholamine drive with maintenance of cerebral perfusion pressure. The maximum tolerated dosing schedule for esmolol can then be tested for patient benefit in subsequent randomized controlled trials.Trial registration ISRCTN, ISRCTN11038397, registered retrospectively 07/01/2021 https://www.isrctn.com/ISRCTN11038397.

Asunto(s)

RESUMEN

OBJECTIVES: Prognostic and diagnostic models must work in their intended clinical setting, proven via "external evaluation", preferably by authors uninvolved with model development. By systematic review, we determined the proportion of models published in high-impact radiological journals that are evaluated subsequently. METHODS: We hand-searched three radiological journals for multivariable diagnostic/prognostic models 2013-2015 inclusive, developed using regression. We assessed completeness of data presentation to allow subsequent external evaluation. We then searched literature to August 2022 to identify external evaluations of these index models. RESULTS: We identified 98 index studies (73 prognostic; 25 diagnostic) describing 145 models. Only 15 (15%) index studies presented an evaluation (two external). No model was updated. Only 20 (20%) studies presented a model equation. Just 7 (15%) studies developing Cox models presented a risk table, and just 4 (9%) presented the baseline hazard. Two (4%) studies developing non-Cox models presented the intercept. Just 20 (20%) articles presented a Kaplan-Meier curve of the final model. The 98 index studies attracted 4224 citations (including 559 self-citations), median 28 per study. We identified just six (6%) subsequent external evaluations of an index model, five of which were external evaluations by researchers uninvolved with model development, and from a different institution. CONCLUSIONS: Very few prognostic or diagnostic models published in radiological literature are evaluated externally, suggesting wasted research effort and resources. Authors' published models should present data sufficient to allow external evaluation by others. To achieve clinical utility, researchers should concentrate on model evaluation and updating rather than continual redevelopment. CLINICAL RELEVANCE STATEMENT: The large majority of prognostic and diagnostic models published in high-impact radiological journals are never evaluated. It would be more efficient for researchers to evaluate existing models rather than practice continual redevelopment. KEY POINTS: • Systematic review of highly cited radiological literature identified few diagnostic or prognostic models that were evaluated subsequently by researchers uninvolved with the original model. • Published radiological models frequently omit important information necessary for others to perform an external evaluation: Only 20% of studies presented a model equation or nomogram. • A large proportion of research citing published models focuses on redevelopment and ignores evaluation and updating, which would be a more efficient use of research resources.

Asunto(s)

RESUMEN

Objective To establish a clinical prediction model for preeclampsia by monitoring risk rating of MP gestation and levels of placental growth factor(PLGF)combined with uterine artery pulsatility index(PI)measured during examination of fetal nuchal translucency(NT).Methods Twenty-four patients with preeclampsia who met the inclusion criteria were selected as the case group,and 95 healthy pregnant women during the same period were randomly selected as the control group.Serum concentrations of PLGF,uterine artery PI values measured by quantitative immunofluorescence assay at 11-14 weeks of gestation,risk ratings for MP hypertension monitoring at 11-20 weeks of gestation,and other relevant data,BMI,age,gestation,mode of delivery,neonatal birth weight and Apgar score were collected in the two groups.Results Results of univariate regression analysis showed that BMI,age,high risk of PI,MP and PLGF<12 were influencing factors for adverse outcomes.Results of multivariate regression analysis showed that high PI,medium high risk in MP and PLGF<12 were independent risk factors for adverse outcomes.The prediction model of PE established was logit(P)=-15.767 + 0.020×PI + 0.072×MP risk(medium-high risk = 1,low risk = 0)+ 0.181×PLGF classification(＜12 = 1,≥12 = 0),with an AUC area of 0.883,specificity of 0.816 and sensitivity of 0.846.Conclusion The combination of PI,MP risk and PLGF to establish a clinical predictive model for preeclampsia has certain value,and its combined predictive value is higher than that of single application.

RESUMEN

Objective@#To explore the performance of machine learning prediction models in forecasting student absenteeism due to respiratory symptoms caused by air pollution in short term, aiming to provide a methodological reference for early warning systems of school diseases.@*Methods@#Utilizing data from shortterm sequences of student absenteeism due to respiratory symptoms in Jiangsu Province from September 2019 to October 2022, the study integrated average concentrations of atmospheric pollutants. A univariate distributed lag nonlinear model was employed to select optimal lag variables for the pollutants. An extreme gradient boosting(XGBoost) algorithm model was developed to predict the frequency of absenteeism due to respiratory symptoms and compared with the seasonal autoregressive integrated moving average with exogenous factors(SARIMAX) model.@*Results@#Between 2019 and 2022, an average of 9 709 students per day in Jiangsu Province were absent due to respiratory symptoms. The daily average air quality index (AQI) was 76.96,with mass concentrations of PM2.5, PM10, NO2, and O3 averaging at 35.75, 61.13, 28.89, 104.81 μg/m3, respectively. Granger causality tests indicated that AQI, PM2.5, PM10, NO2, and O3 were significant predictors of absenteeism frequency due to respirutory symptoms(F=1.46,1.79,1.67,3.41,2.18,P<0.01). The singleday lag effects of PM2.5, PM10, NO2, and O3 reached their peak relative risk (RR) values at lag4, lag0, lag0, lag4 respectively. When integrating these optimal lag variables for the pollutants, the XGBoost model demonstrated superior predictive performance to the SARIMAX model, reducing the mean absolute error (MAE) from 2.251 to 0.475, mean absolute percentage error (MAPE) from 0.429 to 0.080, and root mean square error (RMSE) from 2.582 to 0.713; at the P75 percentile alert threshold, the sensitivity improved from 0.086 to 0.694 and specificity from 0.979 to 0.988, with the Youden index increasing from 0.065 to 0.682.@*Conclusions@#The XGBoost model exhibits robust predictive performance and effective early warning capabilities for shortterm sequences of student absenteeism due to respiratory symptoms caused by air pollution. Schools could timely adopt this model to preemptively detect and control disease outbreaks, thereby enhancing school health management.

RESUMEN

Objective@#To construct a nonsuicidal selfinjury (NSSI) risk prediction model for middle school students using different machine learning algorithms and evaluate the models effectiveness, so as to provide guidance for the prevention and control of NSSI in campus.@*Methods@#In March 2023, a total of 3 372 middle and high school students from schools in Nanchang, Fuzhou and Shangrao cities in Jiangxi Province were selected by combining stratified random cluster sampling and convenient sampling methods. Questionnaire surveys were conducted using various instruments including general information questionnaire, Selfesteem Scale, Ottawa Selfinjury Scale, Social Support Assessment Scale, Chinese Version of the Olweus Bullying Questionnaire, Event Attribution Style Scale, Adolescent Resilience Scale, and Adolescent Life Events Scale. Data were divided into training set (n=2 361) and test set (n=1 011) at a ratio of 7∶3, and variables were selected based on univariate and LASSO regression results. Four machine learning algorithms including namely random forest, support vector machine, Logistic regression and XGBoost, were used to construct NSSI risk prediction models, and the models performance was evaluated and compared using metrics including area under curve (AUC), sensitivity, specificity, positive predictive value, negative predictive value and F1 score.@*Results@#The detection rate of NSSI among middle school students was 34.4%. Univariate analysis showed that there were statistically significant differences in NSSI detection rates among middle school students of different grades, genders, registered residence locations, whether they were class cadres and four types of bullying (physical, verbal, relational bullying and cyberbullying) (χ2=27.17, 15.81, 11.54, 4.63；68.22, 140.63, 77.81, 13.95, P<0.05). NSSI was included as the dependent variable in the LASSO regression model for variable screening, and the results regression identified 10 predictive variables including grade level, selfesteem, subjective support, support utilization, verbal bullying, emotional control, interpersonal relationships, punishment, loss of relatives and property, and health and adaptation issues. The AUC values of random forest, support vector machine, Logistic regression, and XGBoost algorithms were 0.76, 0.76, 0.76 and 0.77, respectively, with no statistically significant differences between pairwise comparisons (Z=-0.59-0.82, P>0.05). Sensitivity values were 0.62, 0.61, 0.62 and 0.61, respectively. Specificity values were 0.74, 0.78, 0.78 and 0.78, respectively. Positive predictive values were 0.56, 0.59, 0.60 and 0.59, respectively. Negative predictive values were 0.79, 0.79, 0.80 and 0.79, respectively. F1 scores were 0.59, 0.60, 0.61 and 0.60, respectively.@*Conclusions@#All four nonsuicidal selfinjury risk prediction models perform well, with the Logistic regression model slightly outperforming the others. Schools and parents should pay attention to the predictive factors corresponding to NSSI, so as to reduce the occurrence of NSSI among middle school students.

RESUMEN

Abstract@#Myopia has become a major public health issue of global concern. Scientific and effective myopia prediction models can help identify high risk groups for myopia, thereby achieving precise prevention. With the rapid development of genome wide association studies and the establishment of large scale prospective population cohorts, the polygenic risk score (PRS) model has been used to predict myopia phenotypes, advancing the myopia prediction window and thus predicting high myopia risk for early screening and intervention for at risk groups. The review aims to systematically elaborate the identification and verification of myopia genes in recent years, briefly describe the practice and effectiveness evaluation of the PRS model in myopia prevention research at home and abroad, reveal the application value in myopia prediction research, and emphasize the relationship between the PRS prediction model and outdoor activities. Close eye use and other preventive measures are of great significance to promote the precise prevention of myopia in children and adolescents.

RESUMEN

ABSTRACT Purpose: The emergency medical service is a fundamental part of healthcare, albeit crowded emergency rooms lead to delayed and low-quality assistance in actual urgent cases. Machine-learning algorithms can provide a smart and effective estimation of emergency patients' volume, which was previously restricted to artificial intelligence (AI) experts in coding and computer science but is now feasible by anyone without any coding experience through auto machine learning. This study aimed to create a machine-learning model designed by an ophthalmologist without any coding experience using AutoML to predict the influx in the emergency department and trauma cases. Methods: A dataset of 356,611 visits at Hospital da Universidade Federal de São Paulo from January 01, 2014 to December 31, 2019 was included in the model training, which included visits/day and the international classification disease code. The training and prediction were made with the Amazon Forecast by 2 ophthalmologists with no prior coding experience. Results: The forecast period predicted a mean emergency patient volume of 216.27/day in p90, 180.75/day in p50, and 140.35/day in p10, and a mean of 7.42 trauma cases/ day in p90, 3.99/day in p50, and 0.56/day in p10. In January of 2020, there were a total of 6,604 patient visits and a mean of 206.37 patients/day, which is 13.5% less than the p50 prediction. This period involved a total of 199 trauma cases and a mean of 6.21 cases/day, which is 55.77% more traumas than that by the p50 prediction. Conclusions: The development of models was previously restricted to data scientists' experts in coding and computer science, but transfer learning autoML has enabled AI development by any person with no code experience mandatory. This study model showed a close value to the actual 2020 January visits, and the only factors that may have influenced the results between the two approaches are holidays and dataset size. This is the first study to apply AutoML in hospital visits forecast, showing a close prediction of the actual hospital influx.

RESUMO Objetivo: Esse estudo tem como objetivo criar um modelo de Machine Learning por um oftalmologista sem experiência em programação utilizando auto Machine Learning predizendo influxo de pacientes em serviço de emergência e casos de trauma. Métodos: Um dataset de 366,610 visitas em Hospital Universitário da Universidade Federal de São Paulo de 01 de janeiro de 2014 até 31 de dezembro de 2019 foi incluído no treinamento do modelo, incluindo visitas/dia e código internacional de doenças. O treinamento e predição foram realizados com o Amazon Forecast por dois oftalmologistas sem experiência com programação. Resultados: O período de previsão estimou um volume de 206,37 pacientes/dia em p90, 180,75 em p50, 140,35 em p10 e média de 7,42 casos de trauma/dia em p90, 3,99 em p50 e 0,56 em p10. Janeiro de 2020 teve um total de 6.604 pacientes e média de 206,37 pacientes/dia, 13,5% menos do que a predição em p50. O período teve um total de 199 casos de trauma e média de 6,21 casos/dia, 55,77% mais casos do que a predição em p50. Conclusão: O desenvolvimento de modelos era restrito a cientistas de dados com experiencia em programação, porém a transferência de ensino com a tecnologia de auto Machine Learning permite o desenvolvimento de algoritmos por qualquer pessoa sem experiencia em programação. Esse estudo mostra um modelo com valores preditos próximos ao que ocorreram em janeiro de 2020. Fatores que podem ter influenciados no resultado foram feriados e tamanho do banco de dados. Esse é o primeiro estudo que aplicada auto Machine Learning em predição de visitas hospitalares com resultados próximos aos que ocorreram.