Predicting Hospital Readmission in Heart Failure Patients in Iran: A Comparison of Various Machine Learning Methods / 대한의료정보학회지

Objectives@#Heart failure (HF) is a common disease with a high hospital readmission rate. This study considered class imbalance and missing data, which are two common issues in medical data. The current study’s main goal was to compare the performance of six machine learning (ML) methods for predicting hospital readmission in HF patients. @*Methods@#In this retrospective cohort study, information of 1,856 HF patients was analyzed. These patients were hospitalized in Farshchian Heart Center in Hamadan Province in Western Iran, from October 2015 to July 2019. The support vector machine (SVM), least-square SVM (LS-SVM), bagging, random forest (RF), AdaBoost, and naïve Bayes (NB) methods were used to predict hospital readmission. These methods’ performance was evaluated using sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Two imputation methods were also used to deal with missing data. @*Results@#Of the 1,856 HF patients, 29.9% had at least one hospital readmission. Among the ML methods, LS-SVM performed the worst, with accuracy in the range of 0.57–0.60, while RF performed the best, with the highest accuracy (range, 0.90–0.91). Other ML methods showed relatively good performance, with accuracy exceeding 0.84 in the test datasets. Furthermore, the performance of the SVM and LS-SVM methods in terms of accuracy was higher with the multiple imputation method than with the median imputation method. @*Conclusions@#This study showed that RF performed better, in terms of accuracy, than other methods for predicting hospital readmission in HF patients.

Factors associated with mortality from tuberculosis in Iran: an application of a generalized estimating equation-based zero-inflated negative binomial model to national registry data / 한국역학회지

OBJECTIVES: Tuberculosis (TB) is a global public health problem that causes morbidity and mortality in millions of people per year. The purpose of this study was to examine the relationship of potential risk factors with TB mortality in Iran. METHODS: This cross-sectional study was performed on 9,151 patients with TB from March 2017 to March 2018 in Iran. Data were gathered from all 429 counties of Iran by the Ministry of Health and Medical Education and Statistical Center of Iran. In this study, a generalized estimating equation-based zero-inflated negative binomial model was used to determine the effect of related factors on TB mortality at the community level. For data analysis, R version 3.4.2 was used with the relevant packages. RESULTS: The risk of mortality from TB was found to increase with the unemployment rate (β

Factors associated with mortality from tuberculosis in Iran: an application of a generalized estimating equation-based zero-inflated negative binomial model to national registry data / 한국역학회지

OBJECTIVES: Tuberculosis (TB) is a global public health problem that causes morbidity and mortality in millions of people per year. The purpose of this study was to examine the relationship of potential risk factors with TB mortality in Iran.METHODS: This cross-sectional study was performed on 9,151 patients with TB from March 2017 to March 2018 in Iran. Data were gathered from all 429 counties of Iran by the Ministry of Health and Medical Education and Statistical Center of Iran. In this study, a generalized estimating equation-based zero-inflated negative binomial model was used to determine the effect of related factors on TB mortality at the community level. For data analysis, R version 3.4.2 was used with the relevant packages.RESULTS: The risk of mortality from TB was found to increase with the unemployment rate (β

Factors associated with mortality from tuberculosis in Iran: an application of a generalized estimating equation-based zero-inflated negative binomial model to national registry data / 한국역학회지

OBJECTIVES@#Tuberculosis (TB) is a global public health problem that causes morbidity and mortality in millions of people per year. The purpose of this study was to examine the relationship of potential risk factors with TB mortality in Iran.@*METHODS@#This cross-sectional study was performed on 9,151 patients with TB from March 2017 to March 2018 in Iran. Data were gathered from all 429 counties of Iran by the Ministry of Health and Medical Education and Statistical Center of Iran. In this study, a generalized estimating equation-based zero-inflated negative binomial model was used to determine the effect of related factors on TB mortality at the community level. For data analysis, R version 3.4.2 was used with the relevant packages.@*RESULTS@#The risk of mortality from TB was found to increase with the unemployment rate (β

Evaluating the impact of risk factors on birth weight and gestational age: a multilevel joint modeling approach

Background: Abnormalities in birth weight and gestational age cause several adverse maternal and infant out- comes. Our study aims to determine the potential factors that affect birth weight and gestational age, and their association

Materials and Methods: We conducted this cross-sectional study of 4415 pregnant women in Tehran, Iran, from July 6-21, 2015. Joint multilevel multiple logistic regression was used in the analysis with demographic and obstetrical variables at the first level, and the hospitals at the second level

Results: We observed the following prevalence rates: preterm [5.5%], term [94%], and postterm [0.5%]. Low birth weight [LBW] had a prevalence rate of 4.8%, whereas the prevalence rate for normal weight was 92.4, and 2.8% for macrosomia. Compared to term, older mother's age [odds ratio [OR]=1.04, 95% confidence interval [CI]: 1.02-1.07], preeclampsia [OR=4.14, 95% CI: 2.71-6.31], multiple pregnancy [OR=18.04, 95% CI: 9.75- 33.38], and use of assisted reproductive technology [ART] [OR=2.47, 95% CI: 1.64-33.73] were associated with preterm birth. Better socioeconomic status [SES] was responsible for decreased odds for postterm birth com- pared to term birth [OR=0.53, 95% CI: 0.37-0.74]. Cases with higher maternal body mass index [BMI] were 1.02 times more likely for macrosomia [95% CI: 1.01-1.04], and male infant sex [OR=1.78, 95% CI: 1.21-2.60]. LBW was related to multiparity [OR=0.59, 95% CI: 0.42-0.82], multiple pregnancy [OR=17.35, 95% CI: 9.73-30.94], and preeclampsia [OR=3.36, 95% CI: 2.15-5.24]

Conclusion: Maternal age, SES, preeclampsia, multiple pregnancy, ART, higher maternal BMI, parity, and male infant sex were determined to be predictive variables for birth weight and gestational age after taking into consideration their association by using a joint multilevel multiple logistic regression model

Diabetic peripheral neuropathy class prediction by multicategory support vector machine model: a cross-sectional study / 한국역학회지

OBJECTIVES: Diabetes is increasing in worldwide prevalence, toward epidemic levels. Diabetic neuropathy, one of the most common complications of diabetes mellitus, is a serious condition that can lead to amputation. This study used a multicategory support vector machine (MSVM) to predict diabetic peripheral neuropathy severity classified into four categories using patients' demographic characteristics and clinical features. METHODS: In this study, the data were collected at the Diabetes Center of Hamadan in Iran. Patients were enrolled by the convenience sampling method. Six hundred patients were recruited. After obtaining informed consent, a questionnaire collecting general information and a neuropathy disability score (NDS) questionnaire were administered. The NDS was used to classify the severity of the disease. We used MSVM with both one-against-all and one-against-one methods and three kernel functions, radial basis function (RBF), linear, and polynomial, to predict the class of disease with an unbalanced dataset. The synthetic minority class oversampling technique algorithm was used to improve model performance. To compare the performance of the models, the mean of accuracy was used. RESULTS: For predicting diabetic neuropathy, a classifier built from a balanced dataset and the RBF kernel function with a one-against-one strategy predicted the class to which a patient belonged with about 76% accuracy. CONCLUSIONS: The results of this study indicate that, in terms of overall classification accuracy, the MSVM model based on a balanced dataset can be useful for predicting the severity of diabetic neuropathy, and it should be further investigated for the prediction of other diseases.

Simultaneous analysis of multidrug resistance 1[MDR1] C3435T, G2677T/A, and C1236T genotypes in Hamadan city population, west of Iran

One of the limitations in the treatment of common diseases such as cancer chemotherapy is development of multidrug resistance [MDR]. Polymorphisms could alter the expression level of MDR1 gene, which plays an important role in MDR. In this research, the frequency of C3435T, C1236T, and G2677T/A polymorphisms of MDR1 gene was investigated in a large group of population from Hamadan city to provide a sample data resource.Peripheral blood [2 ml] was taken, and DNA extraction was carried out. Multiplexed mutagenically separated PCR, which was followed by polyacrylamide gel electrophoresis and silver staining, was applied to detect the mentioned polymorphisms in 935 individuals. Sequencing performed for confirmation of gel electrophoresis resulted in 10 random cases. In total, alleles and genotypes of 933 persons [776 women and 157 men] were determined. The most frequent alleles of the polymorphisms were: 3435T, C1236, and G2677. The most frequent genotypes were: 3435C/T, 1236C/T, and 2677G/A, and their concurrent presence was also found as the most frequent simultaneous genotypes. There was not any meaningful difference among the prevalence of these genotypes in groups of men and women. Our results were close to those of other studies performed in Iran and compared to the other ethnic groups, which showed more similarity to Asian peoples than Europeans. As an aspect of personalized medicine, it could be used by chemotherapists to improve the routine methods of cancer treatment.

Survival Analysis of Gastric Cancer Patients with Incomplete Data

PURPOSE: Survival analysis of gastric cancer patients requires knowledge about factors that affect survival time. This paper attempted to analyze the survival of patients with incomplete registered data by using imputation methods. MATERIALS AND METHODS: Three missing data imputation methods, including regression, expectation maximization algorithm, and multiple imputation (MI) using Monte Carlo Markov Chain methods, were applied to the data of cancer patients referred to the cancer institute at Imam Khomeini Hospital in Tehran in 2003 to 2008. The data included demographic variables, survival times, and censored variable of 471 patients with gastric cancer. After using imputation methods to account for missing covariate data, the data were analyzed using a Cox regression model and the results were compared. RESULTS: The mean patient survival time after diagnosis was 49.1+/-4.4 months. In the complete case analysis, which used information from 100 of the 471 patients, very wide and uninformative confidence intervals were obtained for the chemotherapy and surgery hazard ratios (HRs). However, after imputation, the maximum confidence interval widths for the chemotherapy and surgery HRs were 8.470 and 0.806, respectively. The minimum width corresponded with MI. Furthermore, the minimum Bayesian and Akaike information criteria values correlated with MI (-821.236 and -827.866, respectively). CONCLUSIONS: Missing value imputation increased the estimate precision and accuracy. In addition, MI yielded better results when compared with the expectation maximization algorithm and regression simple imputation methods.

gene selection method for survival prediction in diffuse large B-cell lymphomas patients using 1D discrete wavelet transform

An important aspect of microarray studies includes the prediction of patient survival based on their gene expression profile. To deal with the high dimensionality of this data, use of a dimension reduction procedure along with the survival prediction model is necessary. This study aimed to present a new method based on wavelet transform for survival relevant gene selection. The data included 2042 gene expression measurements from 40 patients with Diffuse Large B-Cell Lymphomas [DLBCL]. The pre-processing gene expression data is decomposed using third level of the 1D discrete wavelet transform. The detail coefficients at levels 1 and 2 are filtered out and expression data reconstructed using the approximation and detailed coefficients at the third level. All the genes are then scored based on the t score. Then genes with the highest scores are selected. By using forward selection method in Cox regression model, significant genes were identified. The results showed wavelet-based gene selection method presents acceptable survival prediction. Using this method, six significant genes were selected. It was indicated the expression of GENE3359X andGENE3968X decreased the survival time, whereas the expression of GENE967X, GENE3980X, GENE3405X andGENE1813X increased the survival time. Wavelet-based gene selection method is a potentially useful tool for the gene selection from microarray data in the context of survival analysis

Real-Data Comparison of Data Mining Methods in Prediction of Diabetes in Iran / 대한의료정보학회지

OBJECTIVES: Diabetes is one of the most common non-communicable diseases in developing countries. Early screening and diagnosis play an important role in effective prevention strategies. This study compared two traditional classification methods (logistic regression and Fisher linear discriminant analysis) and four machine-learning classifiers (neural networks, support vector machines, fuzzy c-mean, and random forests) to classify persons with and without diabetes. METHODS: The data set used in this study included 6,500 subjects from the Iranian national non-communicable diseases risk factors surveillance obtained through a cross-sectional survey. The obtained sample was based on cluster sampling of the Iran population which was conducted in 2005-2009 to assess the prevalence of major non-communicable disease risk factors. Ten risk factors that are commonly associated with diabetes were selected to compare the performance of six classifiers in terms of sensitivity, specificity, total accuracy, and area under the receiver operating characteristic (ROC) curve criteria. RESULTS: Support vector machines showed the highest total accuracy (0.986) as well as area under the ROC (0.979). Also, this method showed high specificity (1.000) and sensitivity (0.820). All other methods produced total accuracy of more than 85%, but for all methods, the sensitivity values were very low (less than 0.350). CONCLUSIONS: The results of this study indicate that, in terms of sensitivity, specificity, and overall classification accuracy, the support vector machine model ranks first among all the classifiers tested in the prediction of diabetes. Therefore, this approach is a promising classifier for predicting diabetes, and it should be further investigated for the prediction of other diseases.

Investigation of qualitative and quantitative of volatile organic compounds of ambient air in the Mahshahr petrochemical complex in 2009

Volatile organic compounds [VOCs] are human-made chemicals widely spread in the environment and produced by petrochemical industries and petroleum refineries. The aim of this research was to evaluate the distribution of VOCs in the ambient air of Mahshahr Petrochemical Complex, Iran. This study was a cross-sectional research performed in 2009. We used the method numbered 1501, 1500, 2000, 1003, 1005, 1010, 2555, 1300 and 1400 of the National Institute of Occupational Safety and Health [NIOSH] for the sampling and analysis of compounds in the air. A total of 204 samples were analyzed using Gas Chroma-tography-Mass Spectrometry [GC-MS] and a Gas Chromatography-Flame Ionization Detector [GC-FID]. The mean of concentrations of the pollutants in the winter is less than in summer and a strong variation occurred among the sampling site, attributed to the change in meteorology. The results indicated high concentrations of benzene in most factories. In addition, a significant difference occurred between the concentrations of the compounds in the ambient air inside and outside the factories in both seasons [P<0.050]. It seems that the atmospheric conditions of the workplace affect the spreading of the pollutants, causing the concentration of the pollutants in the summer to be higher than in the winter. In addition, the frequent prevailing wind speed in the region plays a major role in the distribution of the pollutants from Mahshahr Petrochemical factories

Estimation of the Frequency of Intravenous Drug Users in Hamadan City, Iran, Using the Capture-recapture Method / 한국역학회지

OBJECTIVES: The number of illicit drug users is prone to underestimation. This study aimed to use the capture-recapture method as a statistical procedure for measuring the prevalence of intravenous drug users (IDUs) by estimating the number of unknown IDUs not registered by any of the registry centers. METHODS: This study was conducted in Hamadan City, the west of Iran, in 2012. Three incomplete data sources of IDUs, with partial overlapping data, were assessed including: (a) Volunteer Counseling and Testing Centers (VCTCs); (b) Drop in Centers (DICs); and (c) Outreach Teams (ORTs). A log-linear model was applied for the analysis of three-sample capture-recapture results. Two information criteria were used for model selection including Akaike's Information Criterion and the Bayesian Information Criterion. RESULTS: Out of 1,478 IDUs registered by three centers, 48% were identified by VCTCs, 32% by DICs, and 20% by ORTs. After exclusion of duplicates, 1,369 IDUs remained. According to our findings, there were 9,964 (95% CI, 6,088 to 17,636) IDUs not identified by any of the centers. Hence, the real number of IDUs is expected to be 11,333. Based on these findings, the overall completeness of the three data sources was around 12% (95% CI, 7% to 18%). CONCLUSION: There was a considerable number of IDUs not identified by any of the centers. Although the capture-recapture method is a useful and practical approach for estimating unknown populations, due to the assumptions and limitations of the method, the results must be interpreted with caution.

Survival analysis of breast cancer patients using cox and frailty models

Cox proportional hazard [CPH] model is the most widely used model for survival analysis. When there are unobserved/unmeasured individuals factor, then the results of the Cox proportional hazard model may not be reliable. The purpose of this study was to compare the results of CPH and frailty models in breast cancer [BC] patients. A historical cohort study was carried out using medical records gathered from the Fars Province Cancer Registry. The dataset consisted of 769 women having BC referred to Shiraz Namazi Hospital, south of Iran. These patients had been followed for 6 years. After selecting the most important prognostic risk factors on survival, CPH and gamma-frailty Cox models were used to estimate the effects of the risk factors. The results of CPH model showed that, tumor characteristics and number of involved lymph nodes increase the mortality hazard of BC [P < 0.05]. In addition, the frailty model showed that there is at least a latent factor in the model [P = 0.005]. Both of the frailty and CPH model emphasis that the early detection of BC improves survival in BC patients.

Comparison of ergonomic risk assessment outputs from rapid entire body assessment and quick exposure check in an engine oil company

During the last decades, to assess the risk factors of work-related musculoskeletal disorders [WMSDs], enormous observational methods have been developed. Rapid Entire Body Assessment [REBA] and Quick Exposure Check [QEC] are two general methods in this field. This study aimed to compare ergonomic risk assessment outputs from QEC and REBA in terms of agreement in distribution of postural loading scores based on analysis of working postures. This cross-sectional study was conducted in an engine oil company in which 40 jobs were studied. All jobs were observed by a trained occupational health practitioner. Job information was collected to ensure the completion of ergonomic risk assessment tools, including QEC, and REBA. The result revealed that there was a significant correlation between final scores [r=0.731] and the action levels [r =0.893] of two applied methods. Comparison between the action levels and final scores of two methods showed that there was no significant difference among working departments. Most of studied postures acquired low and moderate risk level in QEC assessment [low risk=20%, moderate risk=50% and High risk=30%] and in REBA assessment [low risk=15%, moderate risk=60% and high risk=25%]. There is a significant correlation between two methods. They have a strong correlation in identifying risky jobs, and determining the potential risk for incidence of WMSDs. Therefore, there is possibility for researchers to apply interchangeably both methods, for postural risk assessment in appropriate working environments

Effects of road deicing salt on the quality of ground water resources in Hamadan province, west of Iran

The aim of this study was to assess the effects of road deicing salt on the quality of the ground water resources in Hamadan Province during winter season. Water samples were taken monthly from thirty wells located around the Hamadan-Asadabad highway. The quality of well water was examined by measuring amount of sodium, chloride, total hardness, total dissolved solid, electrical conductivity, total fecal coliform, and total coliform in well water sample. The correlation between mineral deposits in the water samples and the distance of wells from the highway was investigated using Pearson Correlation Coefficient. It was estimated that nearly 11,000 tons salt were applied annually in this province for deicing roads and streets. There was a statistically significant negative correlation between the quality variables of well water taken from a distance less than 400 meters from highway axis in the southern side except for fecal coliform and total coliform. No statistically significant correlation was seen between the distance from the highway axis and the quality variables of well water taken from the northern side. There was a significant difference between water quality variables of the wells located in a distance less than 200 meters in the northern side of the highway, with that of the wells located in southern side in the highway [p<0.05]. A positive correlation between road dicing salt and mineral deposits in the ground water resources was indicated. Therefore, regarding the limited water resources in Hamadan Province, constraining application of road deicing salt is recommended

Birth distance influential factors: a multilevel recurrent events approach

The study was developed in order to find a subset of potential factors, which affect birth distance pattern, regarding consideration on correlation of events of birth in a family and correlation within clusters/centers which other studies omit these correlations. Referring to documents that were registered for family in the health care centers on socio-economical zone, we consider the families with at least one successful birth. Data were drawn from four health care centers, which selected via 27-health center in Hamedan City, western Iran, each from a socio-economic zone. It was expected, same socio-economic status family have same specific birth distance and a family follows a specific pattern. The multilevel recurrent approach was conducted to analyze the sample. The sample was 480 families and 1115 birth events occurred in these families. The final step model shows that significant factors on the birth distance time were mothers job [P=0.018]. The random effect of second level [clusters/centers] was significant [P=0.038]. In other words, the socio-economic of family affects on the birth distance patterns. Other potential variables were not significantly affected birth distances and were deleted from the final model. There are many potential factors, which may affect to birth distance, but multilevel recurrent event model has a better fit to data because of frailty and center effects. Application of other model such as Cox and frailty models may result in misleading reports

Application of mixture models for estimating the prevalence of cigarette smoking in Hamadan, Iran

So far, several studies were conducted to estimate the prevalence of cigarette smoking in Iran, but none of them used a statistical model to deal with unobserved smokers. The present study planned to estimate the accurate prevalence of cigarette smoking mixture of truncated Poisson distribution. A cross-sectional study was conducted in Hamadan, west of Iran in 2009, using cluster sampling and 1146 men and women aged >/= 18 years were enrolled. The data collection was done by an expert group of psychologists and sociologists. A truncated mixture Poisson distribution was fitted to the daily number of cigarettes smoked by smokers. The number of components of the mixture model and related mean and weight were specified using Bayesian information criteria. Accordingly, the number of cigarette smokers who answered incorrectly to the relevant question was estimated. To investigate the validity of the results, a simulation study was conducted using CAMCR software. Mixture Poisson distribution with four components was the most appropriate model fitted to the count data. After correction for underestimation, the prevalence rate of cigarette smoking in the population was 20.6%, including 36.2% for men and 3.3% for women. According to the simulation study, the bias of estimated prevalence was about zero and the root mean square error was estimated 2.5. The number of unobserved data can be estimated by fitting model to truncated count data. The mixture of truncated Poisson distribution is particularly useful to estimate population size when the main objective of the study is to investigate negative traits to which the participants may answer incorrectly

[ relationship between job stress and occupational accidents in an automobile manufacturing company]

Automobile manufacturing industries are consider among the main sites where accidents are likely to happen. From a cognitive point of view, stress can cause problems such as poor concentration, absent-mindedness, and hesitancy while making decision. This study examines the relationship between job stress and occupational accidents in an automobile manufacturing company. For this descriptive- analytic study, the psychological stress of the occupational groups was measured using standard job stress questionnaire. Safety Behavior Sampling [SBS] technique was used to collect samples of dangerous behaviors. Data were then analyzed using descriptive statistic, ANOVA, Pearson correlation coefficient and Logistic regression tests. Based on our results 88% of the sample suffered from a high level of occupational stress. There was also a statistically significant correlation between level of stress and number of accidents experienced by an individual. [P<0.05]. Having considered the relationship between occupational stress and work-related accidents, minimizing or eliminating risk factors need to be carried out through developing occupational stress management programs as well as following safety measures through reinforcing principles of safe behavior at all levels of organization

Passive or active smoking, which is more relevant to breast cancer

To examine the risk of breast cancer associated with passive and active smoking and to explore risk heterogeneity among studies. We conducted this study in Iran during the year 2006. Fifteen published studies on smoking and breast cancer met the defined criteria. Pooled odds ratio [OR] estimates for female breast cancer were calculated. The active and passive smokers were compared with women categorized as never regularly exposed to tobacco smoke. The pooled risk estimate for breast cancer associated with passive smoking among non-smokers was 1.38 [95% confidence interval [CI]; 1.16-1.65]. The pooled OR for active smokers was 1.25 [95% CI; 1.11-1.41]. Also, the combined OR for passive and active smokers related to breast cancer was 1.30 [95% CI; 1.17-1.45]. Based on the results of the pooled analysis, it can be concluded both passive and active smoking equally increase the risk of female breast cancer

Skin diseases in kidney transplant recipients

The aim of this study was to evaluate the frequency of skin diseases in kidney transplant recipients. This cross-sectional study was performed on 233 kidney transplant recipients in Ekbatan Hospital of Hamedan in 2004. The patients were examined by a dermatologist and diagnosis was made on the basis of clinical observations. Biopsies and scraping of the lesions were taken whenever necessary. Of the patients, 226 [97%] suffered from one or more skin lesions. The most common lesions were drug related, including hypertrichosis, gingival hyperplasia, acne, and cushingoid feature which were detected in 86.7% of the patients. Also, infectious and premalignant or malignant lesions [actinic keratosis, squamous cell carcinoma, and basal cell carcinoma] were seen in 48.9% and 14.2% of the patients. The mean duration of immunosuppressive therapy was significantly higher in patients with infectious skin diseases [P < .001]. Skin lesions are a significant problem in kidney transplant recipients. A careful monitoring of these patients is recommended in order to detect these lesions in early stages and treat them