Hojjat Zeraati, PhD.
Vol 2 No 2 (2016)
Background & Aim: Esophageal cancer is one of the main common causes of death. The high prevalence of esophageal cancer in northern Iran is an important public health problem. The main aim of this paper was to assess the factors affecting survival of atients with esophageal cancer in neighbor provinces around the Caspian Sea using Weibull mixture cure model and mixture cure model based on a non-proportional hazard.
Methods & Materials: This prospective study was designed to gather data of esophageal cancer from the Babol cancer registry, Iran, registered during 1990 to 1991. The study cases were also followed for a period of 15 years up to 2006. Mixture cure model via non-proportional hazard modeling was used to calculate cure fraction and investigate the factors responsible for the cure probability of patients. Estimates were obtained by maximization of the likelihood via SAS proc NLMIXED.
Results: The median survival time was about 9 months and survival probability in 1, 3, and 5 years following diagnosis were 23%, 15% and 13%, respectively. The family history affected the cured fraction independently of its effect on the early outcome In addition, it had significant effect on the probability of uncured state in the both models.
Conclusion: The results demonstrated the great potential in cure modeling survival data via nonproportional hazard model compared to Weibull mixture cure model.
Background & Aim: Nowadays tuberculosis (TB) is one of the public health concerns in Iran. The present study aimed to examine the clinical epidemiology and treatment findings of tuberculosis in Babol, Northern Iran.
Methods & Materials: This cross-sectional study was carried out on medical records of tuberculosis registry pertained to Health Center of Babol, Mazandaran province, Iran. The investigated variables included demographic characteristics and some clinical patterns. To present the findings the descriptive statistics such as mean (standard deviation, SD) and frequency (and relative frequency) were used. Analytical statistics was applied by using chi-square test and independent samples Student’s t-test.
Results: The average age of 200 patients infected with tuberculosis was 47.5 years (SD = 21.4), and 58.5% of the cases were male. In both genders, age groups 18-38 years comprised the greatest percentage. During the 5 years study period, 95% of the patients received treatment protocol 1 (including new cases with positive smear, negative smear, and extra-pulmonary). At the end of treatment, 90.5% were improved or treatment period was completed. The most important clinical symptoms in referring the patients to the therapeutic centers were cough (75.1%) and fever (60.9%).
Conclusion: Education and giving information to general population about the most important clinical symptoms of tuberculosis such as fever and cough might be effective in early detection and prevention of Mycobacterium tuberculosis. Then effective treatment might decrease the burden of the disease.
Background & Aim: In the survival data with Long-term survivors the event has not occurred for all the patients despite long-term follow-up, so the survival time for a certain percent is censored at the end of the study. Mixture cure model was introduced by Boag, 1949 for reaching a more efficient analysis of this set of data. Because of some disadvantages of this model non-mixture cure model was introduced by Chen, 1999, which became well-known promotion time cure model. This model was based on the latent variable distribution of N. Non mixture cure models has obtained much attention after the introduction of the latent activating Scheme of Cooner, 2007, in recent decades, and diverse distributions have been introduced for latent variable.
Methods & Materials: In this article, generalized Poisson- inverse Gaussian distribution (GPIG) will be presented for the latent variable of N, and the novel model which is obtained will be utilized in analyzing long-term survival data caused by skin cancer. To estimate the model parameters with Bayesian approach, numerical methods of Monte Carlo Markov chain will be applied. The comparison drawn between the models is on the basis of deviance information criteria (DIC). The model with the least DIC will be selected as the best model.
Results: The introduced model with GPIG, with deviation criterion of 411.775, had best fitness than Poisson and Poisson-inverse Gaussian distribution with deviation criterion of 426.243 and 414.673, respectively.
Conclusion: In the analyzing long-term survivors, to overcome high skewness and over dispersion using distributions that consist of parameters to estimate these statistics may improve the fitness of model. Using distributions which are converted to simpler distributions in special occasions, can be applied as a criterion for comparing other models.
Background & Aim: The objective of our study was to find out the awareness, inclination and desire to apply the biostatistics as a tool in their areas of work by healthcare professionals.
Methods & Materials: A cross-sectional study was done using self-administered, validated questionnaire, among the faculty, resident doctors and internees of medical college, physiotherapy college, dental college and nursing college. Study was done in North Karnataka, India, during 2012-13.
Results: Out of a total of 500 questionnaires that were distributed among participants, only 460 questionnaires were received with a dropout rate of 8%. The response rate was 92%. Most of the respondents believed that the biostatistics is more difficult than other subjects of medical sciences. The total mean perception of knowledge scores was 21.2 ± 4.26 and total mean attitude scores was 33.4 ± 6.78 and were statistically significant (P < 0.05).
Conclusion: The study has brought out the fact that “biostatistics” is a difficult subject, more of mathematics and is best left to experts. However the respondents have felt “biostatistics” an important part of evidence based medicine and a necessary skill for healthcare professional.
Background & Aim: Multistate models are systems of multivariate survival data where individuals move through a series of istinct states following certain paths of possible transitions. Such models provide a relevant tool for studying observations of a continuous time process at arbitrary times. The aim of this study was to model the transitions from a healthy (hypertension free) state to an illness (hypertension) state of a hypertensive patient under treatment.
Methods & Materials: In this article, the application of multistate modeling using hypertension data is demonstrated. Hospital data were obtained for a cohort of 353 patients from Jimma University Hospital, Ethiopia.
Results: Three states of the Markov process are defined based on the WHO guideline of high blood pressure, state 1 (BP < 140/90 mmHg), state 2 (BP ≥ 140/90 mmHg) and state 3 (dropout). The first state is termed as a healthy state, the second an illness state and the third one is an absorbing state. Initially, the state transition intensities and state occupation probabilities are estimated with no covariate. Then, the effect of gender and family history of hypertension on the state transition intensities are evaluated separately and jointly using proportional intensities model.
Conclusion: The study indicates that gender has a significant effect on the transition intensities but not family history ofhypertension.
Background & Aim: We aimed to describe a standard survival analysis, so that we can analyze some factors related to the time of occurrence of different types of reflux (unilateral-left, unilateralright, and bilateral) in children with antenatal hydronephrosis (ANH) and to provide an approach taking competing risks into account.
Methods & Materials: We used data of 193 children that was collected from Pediatric Urology Research Center of Children’s Hospital Medical Center, affiliated to Tehran University of Medical Sciences, Iran. The cause-specific and subdistribution hazard were computed. P < 0.05 was considered as statistically significant. R packages were used for analyzing the data.
Results: Among these infants (36 girls, 157 boys), 117 (68%) cases had bilateral reflux as the event of interest. The variables “Sex” and “Direction of ANH (in bilateral level)” were significantly different (P<0.05), while “Severity of ANH (in moderate level)” and “Number of other kidney diseases beside ANH and vesicoureteral reflux (VUR)” were borderline. The cumulative incidence derived from the competing risks approach was at a lower level of estimate in comparison with the Kaplan-Meier method. The cumulative incidence curve depicted for the bilateral reflux in subgroups of the sex variable, confirmed the effect of sex.
Conclusion: In the competing risks framework, it is inappropriate to use the Cox and Kaplan-Meier methods, which do not take competing risks into account. Multivariate regression model like the subdistribution hazard model besides the cumulative incidence curve are recommended.
Background & Aim: The aim of the study was to define the epidemiological characteristics of most important infectious diseases in Iran in recent decades.
Methods & Materials: This was a situation trend analysis of infectious diseases (vector and water borne disease, and food borne diseases) in recent decades based on data availability. Three significance levels were used for Mann-Kendall test (90%, 95% and 99%).
Results: The morbidities of most studied diseases had decreased in whole of the country. Unlike other diseases, coetaneous leishmaniasis had not followed the deacreasing trend. In terms of location, Khorasan-e-Shomali was followed the increasing pattern for in four out of six disaeses [malaria, leishmaniasis (coetaneous and visceral), and typhoid].
Conclusion: In conclusion, there is a significant decreasing trend of most important infectious diseases in Iran. Nevertheless, climate change is already happening and would influence the diseases trends. Hence, developing and implementing adaptation strate ies should be considered.
Background & Aim: In many medical studies, one data set is used to construct the model, and to test its performance. This approach is prone to over optimization, and leads to statistics with low chance of external validity. Data splitting can be used to create training and test sets but the cost is reduction in power. The aim of this study was to demonstrate the ability of bootstrap aggregating (bagging) in improving performance of classification and regression tree (CART) models.
Methods & Materials: CART was applied on a sample of 404 subjects, to identify the factors that encourage people to change their body shape by cosmetic surgeries. Comparing known status of subjects with predicted group, sensitivity and specificity of models were compared. Firstly, all data was used to construct the tree and to test its performance. Secondly, model was fitted on half of data and tested on the second half. Thirdly, bagging was applied in which we drew 100 bootstrap samples. Using each bootstrap data, a tree was constr cted and its performance was tested on the unselected subjects. Final group prediction for each subject was determined following majority voting.
Results: When the whole data was used the overall accuracy was 59%. In the test data set and bagging, accuracy reduced to 53% and 56%. Corresponding figures in terms of sensitivity were 60%, 52%, and 55%, respectively.
Conclusion: Bagging corrected performance estimates for over optimization. Bagging method produces statistics which has higher chance for external validity.
Hojjat Zeraati, PhD.
|All the work in this journal are licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.|