Comparing The Value Of Different Cut-off On BISAP Score Predicting Severe Acute Pancreatitis Meta Analysis

Objective:Acute pancreatitis is a common acute abdominal disease,5%-10%of the patients develop into severe acute pancreatitis,which be poor prognosis,high mortality^[1].It is necessary to been treat in intensive care unit（ICU）to strengthen monitoring treatment.The main clinical predicted scores are Ranson score,acute physiology and chronic health evaluation-Ⅱ,（APACHE-Ⅱ）,and computed tomography severity index（CTSI）.The Ranson score includes24h and 48h of admission indicators,which cannot achieve early and dynamic prediction;the APACHE-Ⅱscore is the highest clinical diagnostic accuracy,but includes 18 indicators measurement,the operation is cumbersome and complex;the CTSI score needs to rely on the professional knowledge of the imaging,the clinician result obtained lag,the cost is high,cannot achieve timely,multiple evaluation.Simple scores system and indicators are particularly important for judging the severity of the disease and evaluating the prognosis.Bedside Index of Severity in Acute Pancreatitis（BISAP）include Blood Urea Nitrogen,（BUN>25mg/dl）,Impaired Mental Status,Glasgow Coma Scale score,（GCS<15）,Systemic Inflammatory Response-Syndrome,（SIRS）,Age>60,Pleural effusion detected on imaging.Firstly,BISAP was invented to predict AP patient death.Because it includes indicators which accorded to the latest guidelines recommended predictive,it can get result,early fast,simple and suitable for clinical diagnosis and treatment.Many studies at home and abroad focused on the occurrence SAP score prediction,unanimously affirmed the value of diagnostic SAP,its accuracy is seconded to APACHE-Ⅱscore.Now,BISAP score can be used for early and simple prediction of AP severity,the cutoff is not clearly defined.This study reviewed and analyzed the research literature on the SAP value of BISAP diagnosis in the network database.We screened,evaluated,integrated,and compiled the different cut-off diagnostic value to pay attention to its diagnosis and prediction ability.Methods:We searched the Pub Med,EMBASE,Springer link,Cochrane,CBM,VIP,CNKI,Wanfang database,to obtain the relevant documents,the time span from the earlist to February 2020,skimming,intensive reading.The literature of BISAP diagnostic SAP is evaluated according to the diagnostic analysis literature evaluation scale provided by software Revman5.3.We collected the four-grid table,which is BISAP for SAP diagnosis,used stata15 to integrate and evaluate the data included in the literature.Then we analyzed the heterogeneity among the literature,with different cut-off.We made sensitivity analysis,and set subgroup analysis according to the gold standard and the proportion of SAP.With the heterogeneity is allowed,the data of multiple studies were summarized to obtain the results of the summarized diagnosis.Results:A total of 35 studies were included in the literature on BISAP diagnostic role.There is no obvious deviation between the all of articles.The main difference is in patient selection.13 studies were include to discussed BISAP≥2 diagnostic SAP.The overall heterogeneity was large.9 studies included 1504 patient’s indicators,no heterogeneity,sensitivity is 0.85,specificity is 0.74,diagnostic OR is 16,and area under s ROC curve is 0.88.According to BISAP≥3 diagnostic SAP,19 studies had severe heterogeneity.We included 14 studies,included 2475 patients,then got the result follow sensitivity is 0.53,specificity is 0.95,diagnostic OR is 23,area under s ROC curve is 0.86.The sensitivity was 0.89,specificity was 0.72,diagnostic OR is 21 and the area under the s ROC curve was 0.90.In terms of sensitivity,the former is superior to the latter,but significantly less in specific.According to the gold standard and the proportion of SAP,we set subgroup analysis,the overall results were obtained.After removing the literature,the integration results are obtained.According to reference to Atlanta 2012.There were 5 studies BISAP≥2diagnostic SAP,included 804 patients,the sensitivity was 0.89,the specificity was 0.72,diagnostic OR is 21,the area under s ROC curve was 0.90.4 studies with BISAP≥3 points for diagnosis included 434 patients figured out a sensitivity of 0.55 specificity of 0.91,diagnostic OR is 13 and area under the s ROC curve of 0.89.According to the latest guidelines,there is no significant difference in the predictive value of different cut-off values.If the 1992 Atlanta Diagnostic Guidelines were used as diagnostic criteria,the results would not differ.Then,the BISAP was used to diagnose organ dysfunction,including 8studies had 9 results within 3140 patients.The sensitivity was 0.78,specificity is0.75,diagnostic OR is 11 and the area under the s ROC curve is 0.83.There were five studies recorded predicted SAP by cut-off 2 and included 1443 patient outcomes.The results are as follows,sensitivity was 0.76,specificity 0.79diagnostic ratio 12 and area under s ROC curve 0.84.As well as,the result about cut-off 3 were,Four studies,including 1697 patients,resulted in 0.83,0.71,12,0.85.Conclusion:By BISAP≥2 points to diagnose SAP,the rate of missed diagnosis is low,the rate of misdiagnosis is in moderate range.We should raise vigilance the possibility to develop into SAP,and reinforce monitoring treatment.While BISAP≥3 points to predict SAP,the rate of misdiagnosis is lower,missed diagnosis rate increased and diagnosed SAP basically.With the improvement of gold standard,the efficacy of BISAP score has predicted SAP more accuracy.About BISAP≥2 points to predict of organ dysfunction,the sensitivity and specificity are balanced,worthy recommending in clinical,as complement of the Marshall score.