Predicting students' academic progress and related attributes in first-year medical students: an analysis with artificial neural networks and Naïve Bayes.

BACKGROUND: Dropout and poor academic performance are persistent problems in medical schools in emerging economies. Identifying at-risk students early and knowing the factors that contribute to their success would be useful for designing educational interventions. Educational Data Mining (EDM) methods can identify students at risk of poor academic progress and dropping out. The main goal of this study was to use machine learning models, Artificial Neural Networks (ANN) and Naïve Bayes (NB), to identify first year medical students that succeed academically, using sociodemographic data and academic history. METHODS: Data from seven cohorts (2011 to 2017) of admitted medical students to the National Autonomous University of Mexico (UNAM) Faculty of Medicine in Mexico City were analysed. Data from 7,976 students (2011 to 2017 cohorts) of the program were included. Information from admission diagnostic exam results, academic history, sociodemographic characteristics and family environment was used. The main dataset included 48 variables. The study followed the general knowledge discovery process: pre-processing, data analysis, and validation. Artificial Neural Networks (ANN) and Naïve Bayes (NB) models were used for data mining analysis. RESULTS: ANNs models had slightly better performance in accuracy, sensitivity, and specificity. Both models had better sensitivity when classifying regular students and better specificity when classifying irregular students. Of the 25 variables with highest predictive value in the Naïve Bayes model, percentage of correct answers in the diagnostic exam was the best variable. CONCLUSIONS: Both ANN and Naïve Bayes methods can be useful for predicting medical students' academic achievement in an undergraduate program, based on information of their prior knowledge and socio-demographic factors. Although ANN offered slightly superior results, Naïve Bayes made it possible to obtain an in-depth analysis of how the different variables influenced the model. The use of educational data mining techniques and machine learning classification techniques have potential in medical education.

Descontrol glucémico: desafío de salud pública en prepandemia y pandemia.

OBJETIVO: Comparar el control glucémico entre pacientes con diabetes mellitus tipo 2 prepandemia vs pandemia de Covid-19 que acudieron a unidades de salud de primer nivel de atención de la Ciudad de México. Material y métodos. Se analizaron los registros de 23 912 pacientes con diabetes; 78.7% fueron del grupo prepandemia (2016 a 2020) y 21.3% del grupo pandemia (marzo 2020 a julio 2021). Se calcularon medidas de tendencia central y de dispersión, pruebas t de Student y se ajustó un modelo de regresión logística múltiple. RESULTADOS: La mayoría de los pacientes con diabetes fueron mujeres (66.6 y 62.6%) con edad promedio de 59 y 58 años, respectivamente, y con hemoglobina glucosilada (HbA1) final de 7.7 vs el grupo pandemia (8.0). Las variables asociadas con el descontrol glucémico incluyeron periodo, nivel de HbA1, sobrepeso, obesidad, antecedente de padres con diabetes, número de medicamentos y tipo de insulina. CONCLUSIONES: La mayoría de los pacientes con diabetes en ambos grupos tuvieron descontrol glucémico. Los pacientes del grupo pandemia tuvieron mayor descontrol glucémico de HbA1 comparados con los del grupo prepandemia. Después de recibir atención médica en ambos grupos, los pacientes mejoraron su control glucémico.

Health sciences undergraduate students academic track records.

BACKGROUND: Academic track record analysis is essential for evaluating the training of students and the structure of higher education study programs, which allows diagnosing and preventing educational lag and school dropout. OBJECTIVE: To analyze the differences in academic track records of UNAM health sciences undergraduate students from generations 2001 to 2016. MATERIAL AND METHODS: Study of real cohorts; graduation and lag rates were calculated. ANOVA was used to contrast the graduation rates between campuses by undergraduate program and time. To identify critical periods, survival functions were used with Kaplan-Meier's method. RESULTS: The lowest percentages of lag were observed in nursing and medicine students; nursing students had the highest graduation rates, especially at Zaragoza campus; dentistry students had the lowest graduation rates and the highest dropout and lag rates. Women showed higher graduation rates and lower risk of dropout and lag. CONCLUSIONS: Nursing, medicine and psychology undergraduate students at Zaragoza and Iztacala campuses, with modular programs, achieved the highest graduation percentages and the lowest dropout and lag rates.

ANTECEDENTES: El análisis de las trayectorias académicas es fundamental para evaluar la formación de los estudiantes y la estructura de los programas de estudio de educación superior, lo que permite diagnosticar y prevenir el rezago y abandono escolar. OBJETIVO: Analizar las diferencias en las trayectorias académicas de los estudiantes de las licenciaturas en ciencias de la salud de la UNAM de las generaciones 2001 a 2016. MATERIAL Y MÉTODOS: Estudio de cohortes reales; se calcularon tasas de egreso y rezago. Se realizó ANOVA para contrastar el egreso entre planteles por carrera y tiempo. Para identificar los períodos críticos se utilizaron funciones de supervivencia con el método de Kaplan-Meier. RESULTADOS: En las licenciaturas en enfermería y medicina se observaron los menores porcentajes de rezago; enfermería presentó los mayores porcentajes de egreso, sobre todo en la Facultad de Estudios Superiores Zaragoza; odontología mostró los menores índices de egreso y mayores índices de abandono y rezago. Las mujeres mostraron mayor egreso y menor riesgo de abandono y rezago. CONCLUSIONES: Los estudiantes de las licenciaturas en enfermería, medicina y psicología de las facultades de estudios superiores Zaragoza e Iztacala, con programas modulares, alcanzaron los mayores porcentajes de egreso y menores índices de abandono y rezago.

National examination for medical residency admission: academic performance in a high-stakes test and the need for continuing education.

INTRODUCTION: Performance and selection rate of non-newly graduated physicians in a medical residency admission test as an indicator for the need of continuing education. METHODS: A database comprising 153 654 physicians who took a residency admission test in the period 2014-2018 was analysed. Performance and selection rates were assessed in relation to year of graduation and performance in medical school. RESULTS: The whole sample scored at a mean of 62.3 (SD ±8.9; range 1.11-91.11). Examinees who took the test in their year of graduation performed better (66.10) than those who took the test after their year of graduation (61.84); p<0.001.Selection rates differed accordingly; 33.9% for newly graduated physicians compared with 24.8% in those who took the test at least 1 year after graduation; p<0.001. An association between selection test performance and medical school grades was established using Pearson's correlation: r=0.40 for newly graduated physicians and r=0.30 for non-newly graduated physicians. There were statistically significant differences in selection rates for every ranking group of grades in medical school based on the χ2 test (p<0.001). The selection rates are decreased years after graduation even for candidates with high grades in medical school. DISCUSSION: There is an association between performance in a medical residency admission test and academic variables of the candidates: medical school grades and time elapsed from graduation to test taking. The evidence of decrease in retention of medical knowledge since graduation highlights the pertinence of continuing education interventions.

Trayectorias académicas de los estudiantes de licenciaturas de ciencias de la salud / Health sciences undergraduate students academic track records

Resumen Antecedentes: El análisis de las trayectorias académicas es fundamental para evaluar la formación de los estudiantes y la estructura de los programas de estudio de educación superior, lo que permite diagnosticar y prevenir el rezago y abandono escolar. Objetivo: Analizar las diferencias en las trayectorias académicas de los estudiantes de las licenciaturas en ciencias de la salud de la UNAM de las generaciones 2001 a 2016. Material y métodos: Estudio de cohortes reales; se calcularon tasas de egreso y rezago. Se realizó ANOVA para contrastar el egreso entre planteles por carrera y tiempo. Para identificar los períodos críticos se utilizaron funciones de supervivencia con el método de Kaplan-Meier. Resultados: En las licenciaturas en enfermería y medicina se observaron los menores porcentajes de rezago; enfermería presentó los mayores porcentajes de egreso, sobre todo en la Facultad de Estudios Superiores Zaragoza; odontología mostró los menores índices de egreso y mayores índices de abandono y rezago. Las mujeres mostraron mayor egreso y menor riesgo de abandono y rezago. Conclusiones: Los estudiantes de las licenciaturas en enfermería, medicina y psicología de las facultades de estudios superiores Zaragoza e Iztacala, con programas modulares, alcanzaron los mayores porcentajes de egreso y menores índices de abandono y rezago.

Abstract Background: Academic track record analysis is essential for evaluating the training of students and the structure of higher education study programs, which allows diagnosing and preventing educational lag and school dropout. Objective: To analyze the differences in academic track records of UNAM health sciences undergraduate students from generations 2001 to 2016. Material and methods: Study of real cohorts; graduation and lag rates were calculated. ANOVA was used to contrast the graduation rates between campuses by undergraduate program and time. To identify critical periods, survival functions were used with Kaplan-Meier’s method. Results: The lowest percentages of lag were observed in nursing and medicine students; nursing students had the highest graduation rates, especially at Zaragoza campus; dentistry students had the lowest graduation rates and the highest dropout and lag rates. Women showed higher graduation rates and lower risk of dropout and lag. Conclusions: Nursing, medicine and psychology undergraduate students at Zaragoza and Iztacala campuses, with modular programs, achieved the highest graduation percentages and the lowest dropout and lag rates.

Preoperative risk assessment and spirometry is a cost-effective strategy to reduce post-operative complications and mortality in Mexico.

AIM: Combining preoperative spirometry with the Assess Respiratory Risk in Surgical Patients in Catalunia (ARISCAT) risk scale can reduce post-operative complications and improve patient survival. This study aimed to assess the cost-effectiveness of performing spirometry or not in conjunction with the ARISCAT scale, to reduce post-operative complications and improve survival among adult patients undergoing elective surgery in Mexico. METHODS: A cost-effectiveness analysis (CEA) was performed to compare the specific cost and health outcomes associated with the combined use of the ARISCAT scale and preoperative spirometry (Group 1), and the use of the ARISCAT scale without preoperative spirometry (Group 2). The health outcomes evaluated were post-operative complications and survival. The perspective was from the health care provider (Hospital General de México) and direct medical costs were reported in 2019 US dollars. A decision tree with a time horizon of eight months was used for each health outcome and ARISCAT risk level. RESULTS: The combined use of the ARISCAT scale and spirometry is more cost-effective for reducing post-operative complications in the low and moderate-risk levels and is cost-saving in the high-risk level, than use of the ARISCAT scale without spirometry. To improve patient survival, ARISCAT and spirometry are also more cost-effective at the moderate risk level, and cost-saving for high-risk patients, than using the ARISCAT scale alone. CONCLUSIONS: The use of preoperative spirometry among patients with a high ARISCAT risk level was cost-saving, reduced post-operative complications, and improved survival. Our findings indicate an urgent need to implement spirometry as part of preoperative care in Mexico, which is already the standard of care in other countries.

National examination for medical residency admission: academic variables and performance among different schools / Examen Nacional de Aspirantes a Residencias Médicas: tipo de escuela y desempeño obtenido

Abstract: Objective: To identify medical school characteristics associated with performance in a medical residency admission test. Materials and methods: Performance and selection rates according to type of medical school (Student´s t-test, Chi-squared test), accreditation status (Student´s t-test) and geographic regions (Anova) were analyzed from a database comprising 153 654 physicians who took the residency admission test Examen Nacional de Aspirantes a Residencias Médicas (ENARM) in the period 2014-2018. Results: Performance was 62.5% for accredited programs and 61.4% for non-accredited programs (p<0.001); public schools reached 62.3% and private schools 62.2% (p<0.001). Northern regions performed above 63% while South-Southeast at 58.9% (p<0.001). Selection rate was 26.2% for accredited programs and 22.9% for non-accredited (p<0.001); 26.6% for public schools and 23.6% for private schools (p<0.001). North-East and North-West reached 31% while South-Southeast 20.7%. Conclusions: Type of school, accreditation status and geographic region may influence performance and selection rate.

Resumen: Objetivo: Identificar características de las escuelas de medicina asociadas con desempeño en un examen de admisión a residencias. Material y métodos. Utilizando una base de datos con 153 654 registros de aspirantes entre 2014-2018 se analizaron el desempeño y selección en el Examen Nacional de Aspirantes a Residencias Médicas (ENARM) y su relación con tipo de escuela y estatus de acreditación, así como región geográfica. Resultados: El desempeño fue 62.5% para programas acreditados y 61.4% para no acreditados (p<0.001); 62.3% para escuelas públicas y 62.2% para privadas (p<0.001). Las regiones del norte alcanzaron 63% y Sur-Sureste 58.9% (p<0.001). La tasa de selección fue 26.2% para programas acreditados y 22.9% para no acreditados (p<0.001); 26.6% para escuelas públicas y 23.6% para privadas (p<0.001). Las regiones del norte alcanzaron 31% mientras Sur-Sureste 20.7%. Conclusiones: Las características de la escuela de medicina influencian el desempeño en el ENARM.

National examination for medical residency admission: academic variables and performance among different schools.

OBJECTIVE: To identify medical school characteristics associated with performance in a medical residency admission test. MATERIALS AND METHODS: Performance and selection rates according to type of medical school (Student´s t-test, Chi-squared test), accreditation status (Student´s t-test) and geographic regions (Anova) were analyzed from a database comprising 153 654 physicians who took the residency admission test Examen Nacional de Aspirantes a Residencias Médicas (ENARM) in the period 2014-2018. RESULTS: Performance was 62.5% for accredited programs and 61.4% for non-accredited programs (p<0.001); public schools reached 62.3% and private schools 62.2% (p<0.001). Northern regions performed above 63% while South-Southeast at 58.9% (p<0.001). Selection rate was 26.2% for accredited programs and 22.9% for non-accredited (p<0.001); 26.6% for public schools and 23.6% for private schools (p<0.001). North-East and North-West reached 31% while South-Southeast 20.7%. CONCLUSIONS: Type of school, accreditation status and geographic region may influence performance and selection rate.

Specialty choice determinants among Mexican medical students: a cross-sectional study.

BACKGROUND: The choice of medical specialty is related to multiple factors, students' values, and specialty perceptions. Research in this area is needed in low- and middle-income countries, where the alignment of specialty training with national healthcare needs has a complex local interdependency. The study aimed to identify factors that influence specialty choice among medical students. METHODS: Senior students at the National Autonomous University of Mexico (UNAM) Faculty of Medicine answered a questionnaire covering demographics, personal experiences, vocational features, and other factors related to specialty choice. Chi-square tests and factor analyses were performed. RESULTS: The questionnaire was applied to 714 fifth-year students, and 697 provided complete responses (response rate 81%). The instrument Cronbach's alpha was 0.8. The mean age was 24 ± 1 years; 65% were women. Eighty percent of the students wanted to specialize, and 60% had participated in congresses related to the specialty of interest. Only 5% wanted to remain as general practitioners. The majority (80%) wanted to enter a core specialty: internal medicine (29%), general surgery (24%), pediatrics (11%), gynecology and obstetrics (11%) and family medicine (4%). The relevant variables for specialty choice were grouped in three dimensions: personal values that develop and change during undergraduate training, career needs to be satisfied, and perception of specialty characteristics. CONCLUSIONS: Specialty choice of medical students in a middle-income country public university is influenced by the undergraduate experience, the desire to study a subspecialty and other factors (including having skills related to the specialty and type of patients).