Iranian Journal of Epidemiology

Predicting the Occurrence of Preterm Birth and Determining its Risk Factors Individually Using an Interpretable Machine Learning Model

Akbar Biglarian

Background and Objectives: Identifying pregnant women who are at risk of premature birth and determining its risk factors is essential because it affects their health. This study aimed to use an interpretable machine-learning model to predict premature birth.
Methods: In this study, data from 149,350 births in Tehran in 2019 were utilized from the Iranian Mothers and Babies Network (IMaN) dataset. Various factors related to the mother and the fetus, such as the mother's demographic variables and health status, medical history, pregnancy conditions, childbirth, and associated risks, were considered. The machine learning models, including multilayer neural networks, random forest, and XGBoost, were employed to predict the occurrence of preterm birth after data preprocessing. The models were evaluated based on accuracy, sensitivity, specificity, and area under the ROC curve. The Python programming language version 3.10.0 was applied to analyze the data.
Results: About 8.67% of births were premature. The XGBoost algorithm achieved the highest prediction accuracy (90%). According to the model output, multiple births, which account for 46% of pregnant women's births, had the highest importance score. Delivery risk factors had a score of 41%, and other variables, including neurological and mental illness, preeclampsia, and cardiovascular disease, were subsequently ranked in order of importance for this particular individual.
Conclusion: Using an interpretable machine learning method could predict the occurrence of premature birth. Based on risk factors, the interpretable machine learning method can provide personalized preventive recommendations for every pregnant woman, aiming to reduce the risk of preterm birth.

Observations of the Years of Life Lost During COVID-19 Pandemic in Iran

Kiumarss Nasseri

Background and Objectives: Years of life lost (YLL) or “wasted life” is a measure based on early and untimely death based on the expectation of life at the time of birth. The objective of this study is to measure the YLL during the COVID-19 epidemic in Iran and compare it with a similar antecedent period by age, sex, and province.
Methods: Daily reports of the Ministry of Health and Medical Education on COVID-19 cases and attributed death in the country; Weekly statistics of death and birth, by age, sex, and province reported by the National Organization for Civil Registration; and population data from the Statistical Center of Iran were used in this study.
Results: During the COVID-19 (Corona) epidemic a 27 percent increase in crude death rate was observed compared to similar period before epidemic. During the epidemic period, 319,136 extra deaths was recorded of which 45% was registered as COVID-19 death by Ministry of Health and Medical Education. During this period, a total of 4,897,995 years of life were prematurely lost.
Conclusion: Although this study lacks some detailed analysis due to the limitation of the available data and, it provides a clear picture of the health and demographic impacts of this epidemic in Iran and we can use Information presented in this report in planning and advance preparation for control and management of similar significant epidemics in the future.

Direct and Indirect Estimation of the Prevalence of Waterpipe and Cigarette Smoking Among Adults in Arak: An Approach Based on Network Scale-up Method

Rahmatollah Moradzadeh

Background and Objectives: Indirect methods for estimating hidden populations are essential. The present study aimed to assess the prevalence of cigarette and waterpipe consumption in the Arak metropolis, Iran, directly and indirectly through network scale-up.
Methods: This cross-sectional study was conducted on 1,604 participants. Daily and weekly cigarette and waterpipe consumption data were collected to measure the prevalence directly. The indirect network scale-up method was used to estimate the size of cigarette and hookah smokers. A confidence interval of 95% was considered.
Results: About 49.9% of the participants were men. The average age of men was 39.8 and women 38.7 years. In the direct method, the prevalence of smoking at least one cigarette per day during the last year in women and men was 1.8% (1.0-3.0) and 38.3% (34.9-41.9), respectively. The prevalence of using waterpipe at least once a day during the last year was as much as 0.9% (0.03-1.8) and 4.1% (2.8-5.7) for women and men. In the indirect method of network scale-up during the last year, the prevalence of continuous smoking was 4.8% (4.3-5.0) in women and 19.7% (19.6-19.9) in men. The prevalence of waterpipe in women was calculated at 7.8% (7.8-8.0) and 9.8% (9.7-9.9) in men.
Conclusion: Based on the results, the prevalence of waterpipe and cigarette smoking was high in the Arak, especially among young people. It is suggested to the health system policymakers to pay attention to measures related to reducing the prevalence of these two risk factors, especially among young people, in their health plans.

Evaluating the Effectiveness of the COVID-19 Vaccines on Infection Indicators and Severe Outcomes in Hamadan Province (2021-2022)

Farid Najafi

Background and Objectives: During the COVID-19 pandemic, vaccination was crucial in preventing the spread of SARS-CoV-2 and saving numerous lives. Countries implementing COVID-19 vaccination programs have reported significant reductions in cases, ICU admissions, and COVID-19-related deaths. This study aimed to evaluate the effectiveness of vaccines used in Hamadan province, explicitly focusing on their impact on hospitalization and death caused by COVID-19.
Methods: A test-negative case-control design (TND) was conducted involving patients aged 12 and above who were admitted to hospitals in Hamadan province, Iran, and had symptoms of acute respiratory diseases. Data were extracted from hospital and health system databases. Multiple logistic regression analysis was performed to estimate vaccine effectiveness for the first, second, and reminder doses in prevention of hospitalization, and severe outcomes (ICU admission or death).
Results: The study was conducted on 3,702 patients, and the maximum effectiveness of vaccines against hospitalization was 50% for patients who received a booster dose. The effectiveness of the first dose of vaccine on severe outcomes (admission to ICU or death) was estimated as 42%, but the effectiveness of the vaccines in the second and booster doses were not significant.
Conclusion: Despite the predominant use of inactivated virus vaccines and delayed initiation of vaccination in Iran, this study shows the effect of vaccination on reducing hospitalization and improving the outcomes of COVID-19. The use of more effective vaccines at a more appropriate time plays an important role in reducing the burden on health services and preventing further transmission in future epidemics.

Evaluating the Relationship between Lifestyle and COVID-19 Severity in Patients Admitted to Afzalipour Hospital, Kerman, Iran (2020-2021): A Case-Control Study

AliAkbar Haghdoost

Background and Objectives: This study aimed to evaluate the relationship between the COVID-19 severity and lifestyle among hospitalized patients at Afzali Pour Hospital, Kerman, Iran.
Methods: A total of 289 patients with severe COVID-19 infection (with ICU admission or death) and 293 patients with non-severe type (discharged with no need for hospitalization in ICU) were selected in 2020-2021, and their lifestyle was compared in the last year before the hospitalization.
Results: The adjusted odds ratio (AOR) for severe disease was 1.83 (95% CI: 1.24-2.69) in males compared to females, and 4.35 (95% CI:2.20- 8.59) for people older than 60 years compared to age less than 60. The linear effect of average hours of sleeping during a day was 1.21 (95% CI: 1.08-1.36). The ORs of people who had considerable physical activity at work and people with little activity compared to people who mostly sit in a fixed place were 0.35 (95% CI:0.19-0.65) and 0.44 (95% CI:0.23-0.83), respectively. The use of opioids, traditional remedies, and supplements such as vitamin C raised the risk of severe disease, but the use of vitamin D supplements reduced the risk of severe disease, although this was not statistically significant.
Conclusion: It seems that some aspects of lifestyle, such as lack of physical activity, excessive sleep, and consumption of certain substances, such as opioids, might increase the risk of contracting severe and deadly forms of COVID-19.

Frequency of Eggs’ Antibiotic Residues in Iran: A Systematic Review

Alireza Bahonar

Background and Objectives: Antibiotic residues in food, including eggs, are potentially risky to public health. The objective of this systematic review was to evaluate the relative frequency of antibiotic residues in eggs sold in Iran.
Methods: PubMed, Scopus, Web of Science, Google Scholar, MagIran, Scientific Information Database, and IranDoc were searched. Two independent reviewers screened the titles and abstracts based on the inclusion and exclusion criteria. The inclusion criteria were articles written in English or Persian investigating the relative frequency of antibiotic residues in eggs in Iran, and the exclusion criteria were articles without an available full text. Frequency data, publication year, diagnostic test type, and sampling location were extracted from relevant articles.
Results: Among the 217 identified results, 11 articles were included in the systematic review. ELISA (six articles) was the most frequently used diagnostic test. East Azerbaijan province accounted for the most significant number of studies (four articles) based on the geographical distribution of sampling locations. Seven classes of antibiotics and 12 types of antibiotics were estimated in terms of antibiotic residues, with tetracyclines (5 articles, 6 assessments) having the highest number of assessments. In addition, the highest reported relative frequencies were related to tetracycline residues (100%) in Isfahan and chloramphenicol (75%) in Tabriz.
Conclusion: Tetracycline and chloramphenicol residues had the most significant relative frequency in eggs across Iran.

Utilizing Machine Learning to Predict Antimicrobial Resistance in Bacteria

Monireh Rahimkhani

Antibiotic resistance has increased significantly in recent years. On the other hand, machine learning (ML) algorithms are increasingly used in medical research and healthcare and are gradually improving clinical performance.
Using ML to fight antimicrobial resistance (AMR) is one of the most critical areas of interest among the various applications of these new methods. The rise of antibiotic resistance and managing multidrug-resistant infections that are difficult to treat are important challenges.
Both supervised and unsupervised machine learning tools have been successfully used to predict early antibiotic resistance and thus support clinicians in selecting the appropriate treatment. Machine learning and artificial intelligence (AI) in predicting antimicrobial resistance are among today's sciences. Therefore, an antimicrobial stewardship program (ASP) should be implemented to optimize antibiotic prescribing and limit AMR.