Iranian Journal of

---

Background & Objectives: Although Cox regression is commonly used to detect relationships between patient survival and demographic/clinical variables, there are situations where parametric models can yield more accurate results. The objective of this study was to compare two survival regression methods, namely Cox regression and parametric models, in patients with gastric carcinoma registered at Taleghani Hospital, Tehran.
Methods: Using data from 746 patients who had received care at Taleghani Hospital from February 2003 through January 2007, we compared survival rates between different patient groups with both parametric methods and Cox regression models. The former group included Weibull, exponential and log-normal regression we used the Akaike Information Criterion (AIC) and standardized parameter estimates to compare the efficiency of various models. All the analyses were performed with the SAS software and the level of significance was set at P< 0.05.
Results: The results showed a significantly higher chance of survival in the following subgroups: those with age at diagnosis < 35 years, lower tumor size and those without metastases (P< 0.05). According to AIC, Cox and exponentials model are similar in multivariate analysis but in univariate analysis parametric models are more efficient than Cox, except in the case of tumor size. Log-normal appears to be the best model.
Conclusions: Cox and exponential models have similar performance in multivariate analysis. However, it seems that there is no single model that performs substantially better than others in univariate analysis. The data strongly supported the log-normal regression among parametric models it can give more precise results and can be used as an alternative for Cox in survival analysis of patients with gastric cancer.

---

Background & Objective: Clinically Chronic Allograft Dysfunction (CAD) is characterized by a progressive decline in Glomerular Filtration Rate (GFR) over time, the pattern of disease progression determined by the five-stage model. In this paper, we used Erlang and Hypo-exponential distributions as Phase- Type distributions to describe hazard of kidney failure at over time in RTR with CAD.
Methods: In a single-center retrospective study, 214 patients with RTR with CAD were investigated at the Emam Hospital of Urmia University of Medical Sciences from 1997 to 2005. Kidney function at each visit assessed with GFR and categorized based on NKF and KCOQI staging system.
Results: The estimated hazard rates of disease progression from stage 1 to 2 , 0.0378 from stage 2 to 3 ,0.04 from stage 3 to 4 , 0.0458 and from stage 4 to 5 0.0541 were respectively based on each expected month . This estimates yield a mean waiting time of disease progression from stage 1 to Kidney failure or dialysis 91.63 month. The 18th, 58th, 118th and 155th months of death-censored graft survival were 0.99, 0.75, 0.25 and 0.10 respectively.
Conclusions: The findings of this study are compatible with hyperfiltration theory in chronic kidney disease and give us more detailed information about the daynamic process of disease which would help to manage it effictevliy.

---

Background & Objectives: The Cox proportional-hazards regression and other parametric models model have achieved widespread use in the analysis of time-to-event data with censoring and covariates. However employing Bayesian method has not been widely used or discussed. The aim of this study was to evaluate the prognostic factors in using Bayesian interval censoring analysis.
Methods: This cohort study was based on 178 patients with gastric cancer from January 2003 to December 2007 admitted to Taleghani teaching hospital in Tehran. Known prognostic risk factors were entered into the analysis using Bayesian Weibull and Exponential models. The term DIC was employed to find best model.
Results: The results were showed survival rate depended on age of diagnosis and tumor size. Those patients who had early diagnosis and/or had smaller tumor size were in lower risk of death.
Conclusion: The age of diagnosis and tumor size of patients are important prognostic factors related to survival of patients with gastric cancer. Based on DIC, Bayesian analysis of the Weilbull model performed better than the Exponential model. As a result, if this cancer has been diagnosed early, the relative risk of death would reduce.

---

Background & Objective: Using parametric models is common approach in survival analysis. In the recent years, artificial neural network (ANN) models have increasingly used in survival prediction. The aim of this study was to predict of survival rate of patients with gastric cancer by using a parametric regression and ANN models and compare these methods.
Methods: We used the data of 436 gastric cancer patients from a cancer registry in Tehran between 2002-2007. All patients had a confirmed diagnosis. Data were randomly divided into two groups: training and testing (or validation) set. For analysis of data we used a parametric model (exponential, Weibull, normal, lognormal, logistic and log-logistic models) and a three layer ANN model. In order to compare of the prediction of two models, we used the area under receiver operating characteristic (AUROC) curve, classification table and concordance index.
Results: The prediction accuracy of the ANN and the parametric (Weibull) models were 79.45% and 73.97% respectively. The AUROC for the ANN and the Weibull models were 0.815 and 0.748 respectively.
Conclusions: The ANN had a better predictions than the Weibull model. Thus it is suggested to use of the ANN model survival prediction in field of cancer.

---

Background & Objectives: Cholera is always being considered as a public health threat in poor and developing countries. However outbreaks of cholera are not very common in central area of Iran in 2008 district health authority reported a cluster of diarrhea cases. We investigated this cluster to identify the etiological agent, source of transmission and propose control measures.
Methods: We analyzed the data of total of 1219 patients with colorectal cancer who registered between 1 January 2002 to 1 October 2007. Data were analyzed using univariate and multivariate Accelerated Failure Time (AFT) parametric survival model with frailty, utilizing STATA statistical software.
Results: In the univariate analysis for age at diagnosis, gender, marital status, race and education level, the survival of patients with colon cancer were approximately between half to one fourth and for BMI, alcohol history, Inflammatory Bowel Disease (IBD), familial history of cancer and the pathologic stage of tumor, the survival of patients with colon cancer were significantly (between 0.12 to 0.56 times) shorter than those patients with rectal cancer. In the multivariate analysis, for age at diagnosis (45-65 years), there was significant difference between colon and rectum cancer. But for BMI, alcohol history, IBD and pathologic stage there were not significant differences. The adjusted survival and 1, 2, 3, 4 and 5 year survival of patients with rectal cancer were better than those with colon cancer.
Conclusions: Site-specific evaluation of colon and rectum could give a better perspective of factors affecting these cancers. It may help to design of clinical trials, better diagnosis of diseases and optimal administration of specific treatments.

---

Background & objectives: Competing risk data is one of the multivarite survival data. Competing risk data can be modelled using copula function. In this study we propose a bayesian modelling approach of competing risk data using the copula function.
Methods: We used the data from colorectal cancer registyrarty in Tehran. After constructing likelihood function using Clayton copula by choosing appropriate prior distribution for parameters, we obtained the posterior distribution of parameters using the Metropolis-Hastings algorithms and Slice sampling.
Results: The results of univariate analysis showed that sex, histology of tumor, extent of wall penetration, lymph node metastasis, distant metastasis and pathological stage of tumor were significantly associated with colon cancer and sex, histology of tumor, lymph node metastasis, distant metastasis and pathological stage of tumor were were significantly related to rectal cancer. In the multivariate analysis, age at diagnosis, tumor grade and distant metastasis were significant prognostic factors for colon cancer and tumor grade and size of the tumor were significant prognostic factors of rectal cancer
Conclusions: As we showed some variables may have different impacts on colon and rectum cancers, consequently, further studies are needed to be conducted considering risk factors of these cancers separately.

---

Background & Objectives: Changing the pattern of mortality gives important perspective of health determinants. The aim of this study is to detect location and time of mortality pattern change in country using statistical change point method during 1971-2009 Years.
Methods: We assume for years before and after 0 k , t y has a Poisson distribution with means 0 l and 1 l , respectively. We used several methods for estimation change point in real data by assume Poisson model.
Results: Using two simulated and real data analysis showed that the change point has been occurred in year 1993 and this confirmed by all methods.
Conclusion: Our findings have shown that the change pattern of mortality trend in Iran is related to improvement of health indicators and decreasing mortality rate in Iran.

---

Background & Objectives: Identifying the important influential factors is a great challenge in oncology studies. Decision tree is one of methods that could be used to evaluate the prognostic factors and classifying the patients' homogeneously. This method identifies the main prognostic factors and then determines the subgroups of patients based on those prognostic factors. The aim of this study was to assess the prognostic factors and homogeneous subgroups of colorectal patient through survival tree.
Methods: Data collected from an observational of 739 colorectal patients registered in the cancer registry affiliated to the center of Research Center of Gastroenterology and Liver Disease (RCGLD), Shahid Beheshti Medical University, Tehran, Iran. Death was the interested event and the survival time was calculated from date of diagnosis until occurrence of event (or censoring) in months. Finally we used decision tree based method for classifying and analyzing the data.
Results: Based on our result, decision tree identified four covariates as important prognostic factors in 0.05 significant levels: general stage of cancer, age of diagnosis, histology of tumor and morphology type of tumor. Also patients based on these prognostic factors divided into five homogeneous subgroups. The greater values of measure of separation (SEP) criterion support the appropriateness of this model for such the data.
Conclusion: Decision tree is powerful and intuitive method. It has a key feature that is in addition to evaluate the prognostic factors, provides the homogeneous subgroups for future analysis.

---

Background & Objectives: Cox regression model is one of the statistical methods in survival analysis. The use of smoothing techniques in Cox model makes the more accurate estimates for the parameters. Fractional polynomial is one of these techniques in Cox model. The aim of this study was to assess the effects of prognostic factors on survival of patients with gastric cancer using the fractional polynomial in Cox model and Cox proportional hazards.
Methods: Information of total of 216 patients with gastric cancer who underwent surgery in the gastroenterology ward of Taleghani Hospital in Tehran between 2003 and 2008 were included in this retrospective study. In this research, fractional polynomial in Cox model and Cox proportional hazards model were utilized for determining the effects of prognostic factors on patients’ survival time with gastric cancer. The SPSS version 18.0 and R version 2.14.1 were used for data analysis. These models were compared with Akaike information criterion.
 Results: The analysis of Cox proportional hazards and fractional polynomial models resulted in age at diagnosis and tumor size as prognostic factors on survival time of patients with gastric cancer independently (P<0.05). Also, Akaike information criterion was equal in both models.
Conclusion: In the present study, the Cox proportional hazards and fractional polynomial models led to similar results with equal Akaike information criterions. Using of smoothing methods helped us eliminate non-linear effects but it seemed more appropriate to use Cox proportional hazards model in medical data because of its’ ease of interpretation and capability of modeling in both continuous and discrete covariates.

---

  Background and Objectives : The aim of this study was to assess the disease trajectory and recurrence rate of pemphigus based on the analysis of the gap time between successive recurrent events. In this regard, the most important associated factors with the risk of recurrence could be explained.

  Methods: This longitudinal study was performed on 112 pemphigus patients who attended the dermatology department of Imam Khomeini Hospital, Tehran, Iran, from March 2006 to January 2013. The study duration was considered from the diagnosis of the disease to December 2013. Recurrent events were analyzed based on the gap time between successive events using the multivariate time dependent frailty model. The time between two recurrent gap times was determined monthly between two successive events.

  Results : Decreasing the gap times between two successive events indicates that the subsequent event after the first recurrence occurs with shorter time intervals. So, the disease trajectory represents an increase in the recurrence rate over time. Based on the results of multivariate frailty model, IgG antibody's level was the only effective factor on the recurrence hazard rate of the patients. Also, this model proved that the frailty effects were time dependent frailties.

  Conclusion: Assessing the disease trajectory and recurrence hazard rate can be achieved through analyzing the gap time between successive recurrent events. This analysis also identifies the factors that influence the risk of subsequent recurrent events.

---

Background and Objectives: One of the main concerns of heart specialists is the occurrence of restenosis after coronary angioplasty which can lead to coronary artery bypass graft, myocardial infarction, and death. The present study was conducted to investigate the factors affecting the frequency of restenosis during four years in patients of Zanjan. 

Methods: In the present retrospective cohort study, all the patients who underwent angioplasty in Ayatollah Musavi Hospital of Zanjan from April of 2009 to June of 2011 were examined in terms of the frequency of restenosis. According to the patients’ medical records, all the demographic and clinical data of the patients were collected. Since the dependent variable was count in nature and the data were over-dispersed, negative binomial regression was used for modeling.

Results: The incidence of at least one restenosis during four years after angioplasty was calculated to be 43%. According to the negative binomial regression model, the ratio of restenosis in patients suffering from diabetes, unstable angina, chronic kidney disease, and myocardial infarction was 32%, 44%, 66%, and 30% more than other patients, respectively (P<0.05).

Conclusion: In the present study, the effective factors of restenosis were recognized as diabetes, unstable angina, chronic kidney disease, and history of myocardial infarction; hence, assessment and periodic follow-up of these patients are strongly recommended.

---

Background and Objectives: In studies in which each person may experience an event at different times, they are recurrent events.One of the most popular approaches in analyzing recurrent event is obtaining an estimate of the means/rate of events at different times. In this context,one of the things that could help to better understand the effect of this factor on the response is determining the variability due to quantitative variables in the rate of events over time. In this study,we applied kernel and B-spline methods to estimate coefficients in the time dependent-coefficients rate model and showed its application in data of psoriasis.

Methods: In this study,data of patients with psoriasis who had a relapse leading to hospitalization in the Dermatology Department of Imam Khomeini Hospital,between 2005 and June 2013 were used. To investigate the relapse rate during these years,time-dependent coefficients rate model was used and the variability of these effects was assessed using the Wald test. Both b-Spline and kernel methods were used for estimating time varying coefficients in the time-dependent rate model.Finally,the results of the methods were compared based on estimates obtained.

Results: The results of this study showed that according to Wald test,the effect of the variables such as the season on the occurrence of psoriasis was significantly different (P-value<0.01).Also, according to the estimated coefficients from both methods,there was a little difference between them.

Conclusion: When the effect of a variable on the occurrence of the events is different at different time, then time-dependent coefficients rate model may provide a better estimate of the effect of variable on response.

---

Background and Objectives: Annually about 7 million premature deaths occur due to air pollution in the world. Nitrogen oxides are among major air pollutants. Although many foreign exposure assessment studies have been carried out, Iranian studies are limited to primary analyses. Hence, in this research, we studied spatial variation of nitrogen oxides using spatiotemporal modeling in Tehran 2014.
 
Methods: The concentration of nitrogen oxides was obtained from 21 air pollution monitoring stations in Tehran. There were 8760 records for each pollutant in each station. Holidays and land elevation were the predictors implemented in the spatiotemporal model. The D-STEM software was used for analyses and mapping.
 
Results: Nitrogen monoxide significantly decreased (P<0.001) over holidays and with an increase in land elevation (coefficient: -0.070 and -0.169, respectively). Moreover, the concentration of nitrogen dioxides decreased in holidays (coefficient: -0.630) but increased with with an increase in land elevation (coefficient: 0.155) (P<0.001).
 
Conclusion: Spatiotemporal exposure assessment of nitrogen oxide pollutants was done for residents of Tehran for the first time in this study. The results of this study, which are estimation maps for daily nitrogen oxides, could benefit future epidemiological studies in evaluation of the effect of air pollutions on health of Tehran citizens.
 

---

Background and Objectives: Academic failure is of paramount importance for medical students because it might lead to a decline in scientific level of the community of physicians in the future. This study was conducted to investigate the predictors of academic failure in medical students of Lorestan University of Medical Sciences using classification tree. 
 
Methods: In this cohort study, academic records of all medical students of Lorestan University of Medical Sciences during the academic years of 1999-2008 were selected by census and were followed up until September 2016. Academic failure was defined as having at least one of the components of appropriate grade point average, prolonged graduation, academic probation, dropout, expulsion, and any failure in ccomprehensive exams and the CART classification tree was adopted using the SPSS 22 software to predict it.
 
Results: The cumulative incidence of academic failure was 26.4% and the most prevalent components were prolonged graduation (21.7%) and academic probation (15.0%). The probability of academic failure was 0.449 in subjects taking guest courses, 0.220 in subjects with no history of guest courses admitted to courses with less than 40 students and admission quotas of zone 1 or 3, and 0.456 in subjects with no history of guest courses admitted to courses with more than 40 students and males.
 
Conclusion: With respect to identifying the predictors of academic failure, it is suggested that these students be referred to consulting centers of the university or educational supervisors’ moreover, the regulations of taking guest courses in other universities should be revised.

---

Background and Objectives: The purpose of this study was to predict the mortality rate of colorectal cancer in Iranian patients and determine the effective factors  on the mortality of patients with colorectal cancer using random forest and logistic regression methods.
 
Methods: Data from 304 patients with colorectal cancer registry from the Gastroenterology and Liver Research Center of Shahid Beheshti University of Medical Sciences during the years 2009 to 2014 were used as a retrospective study. Data analysis was performed using random forest and logistic regression methods. To analyze the data, R software version 3.4.3 was considered.
 
Results: Ten important variables related to colorectal cancer deaths were selected by random forest method. Several criteria such as the area under the characteristic curve (AUC) were used to compare the random forest method with logistic regression. According to both criteria, five important variables ranked by random forest were Cancer stage, age of diagnosis, patient's age, HLA, and degree of differentiation (tumor differentiation). In terms of different criteria, the random forest method had better performance than logistic regression (Area under the ROC curve for random forest and logistic regression methods was: 98%; 80% respectively).
 
Conclusion: Variables such as Cancer stage, age of diagnosis, patient's age, HLA, and degree of differentiation are considered as the most important factors affecting mortality in colorectal cancer, that the patients' longevity can be increased with the early diagnosis of cancer and screening programs.
 

---

Background and Objectives: In many medical situations, people can experience recurrent events with a terminal event. If the terminal event is considered a censor in this type of data, the assumption of independence in the analysis of survival data may be violated. This study was conducted to investigate joint modeling of frequent events and a final event (death) in breast cancer patients using a frailty model.
 
Methods: In this retrospective cohort study, 443 patients with breast cancer registered at the Hospital of Shohadaye Tajrish Cancer Research Center were studied. The model of Liu (2004) was applied for joint modeling of recurrent events and a terminal event in which a shared frailty with gamma-distribution was used. Data modeling and data analysis were done using the R software.
 
Results: Four hundred and forty three women with breast cancer were studied. Univariate and multivariate analysis were performed in these patients. Of these, 338 cases (76.3%) had recurrence events, and 105 (23.7%) were censored. The obtained results of joint frailty model indicated that the relative risk of relapse in patients with a positive first-degree family history was 36% higher than that of other people (P<0.05). The relative risk of relapse in patients with stage 3 disease was 19% more than other stages and also the relative risk of relapse in patients with chemotherapy was 2.5 times higher than those without chemotherapy.
 
Conclusion: In this study, the presented model, in addition to simultaneous modeling capability of the event, could help prevent a higher prevalence of the terminal event (death) and thus reduce the adverse effects of reversible diseases.