Iranian Journal of

---

Background and Objectives: A considerable number of clinical trials are conducted in Iran each year. Not much is known about the characteristics of them, this study aimed to investigate key characteristics of Iranian clinical trials.
 
Methods: All clinical trial protocols registered in IRCT until November 2013 were selected. Text mining techniques were used to extract information from data that was mostly in the textual format. Stata software (v. 11) was used for both extraction and analysis.
 
Results: Totally 5049 trials where analyzed. Trials were in 40 different areas of medicine and their annual number was increasing. The sample size of registered trials was less than 70 in most trials and showed no significant improvement during the 5-year period (Median sample size= 64, inter quartile range (IQR) = 40-100). Phase III trials comprised 8.4% of the trials with a median sample size of 75 (IQR= 50-120). Of all phase III trials, 69% had a target sample size of less than 100. Most trials used randomization techniques (87.2%) but failed to use placebo (65.3%) and double/triple-blinding techniques (55.8%). The leading sponsors for 99.9% of trials were national bodies and 92.2% of the trials received public funds. Only 1% of the trials were exclusively sponsored by pharmaceutical companies.
 
Conclusion: The majority of clinical trials conducted in Iran are investigator-initiated and are funded by the public sector. The small sample size and failures in bias control observed in the majority of trials highlights the need for revising current policies of conducting trials in Iran.

---

Background and Objectives: Gestational Diabetes Mellitus (GDM) is the most common metabolic disorder in pregnancy. In case of early detection, some of its complications can be prevented. The aim of this study was to investigate early prediction of GDM by logistic regression (LR), discriminant analysis (DA), decision tree (DT) and perceptron artificial neural network (ANN) and to compare these models.
 
Methods: The medical files of 420 pregnant women (2010-12) in Kermanshah health centers were evaluated using convenience sampling. Demographic data, pregnancy-related variables, lab tests results, and a diagnosis of GDM according to a fasting blood sugar level of 92 or more were collected from their files. After fitting the four models, the performance of the models was compared and according to the criteria of accuracy, sensitivity and specificity (based on the ROC curve), the superior model was introduced.
 
Results: Following the fitting of LR, DA, DT and perceptron ANN models, the following results were obtained. The accuracy of the above models was 0.81, 0.83, 0.78 and 0.83, respectively, the sensitivity of the models was 0.50, 0.63, 0.58 and 0.58, the specificity of the models was 0.96, 0.93, 0.87 and 0.94, and the area under the ROC curve was 0.86, 0.78, 0.73 and 0.87, respectively.
 
Conclusion: In predicting and categorizing the presence of GDM, the ANN model had a lower error rate and a higher area under the ROC curve compared to other models. It can be concluded that this model offers better predictions and is closer to reality than other models.

---