ABSTRACT
ObjectiveTo analyze the risk factors for false negative diagnosis of obscure gastrointestinal bleeding (OGIB) by capsule endoscopy.MethodsA total of 133 OGIB inpatients,104 in true positive group and 29 in false negativc group,were reviewed.The features of demography,diseases and capsule endoscopy were collected and then analyzed,which included 10 variables like age,sex,time of bleeding,diseases accompanied,type and location of the disease,hemoglobin concentration,transit time of CE,quality of CE pictures and type of purgative agents.All data were analyzed with t test,and all the enumeration data were analyzed with chi square test.Logistic regression was used to analyze the correlation between the factors and results of diagnosis.ResultsAge ( t =2.095,P =0.038 ),concentration of hemoglobin ( t =2.143,P=0.034),type (X2 =20.222,P ＜0.001) and location (X2 =33.732,P ＜0.001) of the diseases,image quality of the CE (X2 =9.219,P =0.002 ) and the type of purgative agents (X2 =6.999,P =0.024) were found to have statistical differences between the two groups.Chi-square and partition Chi-square test revealed the occurrence of civerticulosis and lesion location,i.e.lower ileum and ileumcecum,were of statistical difference between the two groups (X2 =22.233,P ＜ 0.001 and x2 =24.412,P ＜ 0.001 ).Univariate logistic regression showed diverticulosis ( OR =0.102,P ＜0.001 ),lower bowel diseases ( OR =0.110,P ＜0.001 ),poor quality of CE pictures ( OR =0.258,P =0.004 ) and the use of sodium phosphate agent ( OR =0.367,P =0.027) were risk factors for false negative diagnosis,while older age facilitated diagnosis ( OR =1.024,P =0.041 ).However,multivariate logistic regression showed no statistic significance in type of purgative agent ( P =0.05 ) or the concentration of hemoglobin ( P =0.394).Furthermore,elder age facilitated positive diagnosis ( OR =1.031,P =0.032),while diverticulosis ( OR =0.118,P =0.001 ),lower bowel diseases ( OR =0.145,P =0.001 ) and poor quality of CE pictures ( OR =0.245,P =0.016) were correlated with higher probability of false negative diagnosis.ConclusionAge,disease type,disease location and image quality exert great influence on CE diagnosis.Diverticulosis,lower location of the diseases and poor CE image quality are risk factors for false negative diagnosis.
ABSTRACT
ObjectiveTo evaluate the clinical value of capsule endoscopy (CE) image screening system developed by our group.MethodsBetween March 2011 and June 2011,50 patients (29 males and 21 females with a mean age of 45.7(45.7 ± 4.1 ) ) underwent CE in our department.CE images were reviewed,and processed by the screening system at the rate of 50％,60％,70％,80％ and 90％,respectively.After reviewing the remaining and the deleted frames,the endoscopists calculated the remaining rate of the lesions by recording the number of lesions of all patients and the number of remaining lesions.The procedure time and the total number of the frames of each case were also recorded.ResultsA total of 39 patients of 50 had lesions and the detection rate was 78％.91 lesions were detected in the 39 patients,most of whom had more than 2 lesions.When the prescribed deletion rate was less than or equal to 60％,the lesion remaining rate reached 100％,with a mean frames of the cases of 46 242,and the mean processing time of 15.62 min.ConclusionWith all the lesions remained,the system can delete 60％ similar frames in a very short time,which greatly reduces the diagnosis time and workload for endoscopist,and improves the efficiency of capsule endoscopy.