Sensitivity Analysis for Hidden Bias in Uncontrolled Clinical Trials And It Applications for Comparing Two Treatment Regimens of Patients With Nonalcoholic Fatty Liver

randomized controlled trials are gold standard for investigate treatment effect and important characteristics among the groups is random assignment. In studies، there is no random allocation between the groups; hidden bias may be defined results.

We performed sensitivity analysis method was proposed by Mr. Resenbum using R software and programming environments were analyzed. The study population included patients who had been diagnosed with fatty liver by ultrasound.

The analysis of the results with routine statistical tests showed there was statistically significant in liver enzymes (ALT، AST and AUK) between groups. However، sensitivity analysis model showed that results were sensitive to hidden bias and lead to disturbed results.

The results of the model sensitivity analysis shows that vitamin E has no effect on liver enzymes. In this study، the sensitivity is low to hidden bias and has affected the results.