Objective To investigate the etiologies, mechanisms, diagnosis and management of postoperative gastroduodenal fistulas in patients with severe acute pancreatitis (SAP). Methods The clinical data of 18 cases of postoperative gastroduodenal fistulas (PGF) in patients with SAP admitted in our hospital from Jan, 1996 to Dec, 2007 were analyzed retrospectively. Results Of 18 patients with SAP, 4 patients underwent cholecystectomy and common bile duct exploration, 4 patients underwent pancreatic debridement and open saucer-type drainage, 10 patients underwent pancreatic debridement and pancreas-bed mobilization, abdominal closed drainage. Duodenal fistula occurred in 12 patients (66.7%), gastric fistula occurred in 6 patients (33.3%). 14 patients (77.8%) developed gastroduodenal fistula within 3 weeks of surgery, and 4 patients developed gastroduodenal fistula after 3 weeks of surgery. 16 patients (88.9%) were cured with non-surgical management including drainage, antibiotics and abscess cavity douching. The average length of stay was 65 days. Two patients (11.1%) died of intra-abdominal hemorrhage and intra-abdominal infection complicated with abdominal compartment syndrome and multiple organ dysfunction syndromes, respectively. Conclusions SAP complicated with postoperative gaatroduodenal fistulas could be cured by appropriate non-surgical treatment.

OBJECTIVE To excavate the risk signals of adverse drug events （ADEs） of atezolizumab after marketing， so as to provide a reference for rational use of atezolizumab. METHODS Using the FAERS database， the ADEs of atezolizumab were collected from marketing on May 18th 2016 to Mar. 31st 2022. The ADEs risk signals were mined by reported odds ratio （ROR） and medicines and Healthcare Products Regulatory Agency method （MHRA）， and then described and classified by preferred term （PT） and system organ classification （SOC） in the adverse drug reaction glossary of Medical Dictionary for Regulatory Activities （MedDRA） （23.0 edition）. RESULTS A total of 16 051 ADEs reports with atezolizumab as the primary suspected drug were collected， and 401 PT were mined， involving 17 SOC. In 16 051 ADEs reports， the majority of patients were male （51.64%）， with main age distribution of 19 to 64 years old and more than 65 years old. According to the number of reports， ADEs were mainly focusing on pyrexia， disease progression， interstitial lung disease， anaemia， pneumonia， etc. The involved SOC includes kidney and urinary system diseases， benign， malignant and unknown tumors （including cysts and polyps）， endocrine system diseases， etc. According to the signal intensity， the new ADEs not specified in the drug instructions were macular thickening， paraneoplastic neurological syndrome， reactive perforating collagenosis， paraneoplastic dermatomyositis， anti-erythrocyte antibody positive and liver sarcoidosis. CONCLUSIONS In addition to the ADEs mentioned in the drug instructions，when using atezolizumab in clinic，attention should also be closely paid to ADEs of eyes， nervous and skin system， so as to ensure the safety of drug use in patients.

OBJECTIVE To identify Ginseng Flos and their confounds by using the high-throughput sequencing technology, and to verify the accuracy of high-throughput sequencing technology in species identification by using ITS2 sequencing technology. METHODS High-throughput sequencing was performed on the amplified products of Ginseng Flos adulterated samples, use cutadapt, PEAR, PRINSEQ, Usearch, RDP classifier, SINTAX software to obtain operational taxonomic unit(OUT) sequences, remove fungi, unclassified and other non-green plant sequences. To avoid false positives, delete OTU sequences with a sequence number <100 or base numbers <200 bp. The ITS2 amplification products of Ginseng Flos, Quinquefolii Flos, and Notoginseng Flos were sequenced. To verify the accuracy of high-throughput sequencing technology for species identification, MEGA 11.0 was used to construct neighbor joining system cluster tree, genetic distance, interspecific information loci and Blast analysis of ITS2 and OTU base sequences of Ginseng Flos, Quinquefolii Flos, and Notoginseng Flos. RESULTS A total of 54 653 valid sequences were obtained by high-throughput sequencing, the serial numbers of Ginseng Flos, Quinquefolii Flos, and Notoginseng Flos were OTU1, OTU2, OTU3, respectively, and the corresponding effective sequences were 31 325, 857 and 442, respectively. By performing a Blast search of ITS2 and OTU base sequences of each species, each species was supported. The genetic distance between Ginseng Flos and Quinquefolii Flos and Notoginseng Flos was 0.010 and 0.033, respectively. Ginseng Flos and Quinquefolii Flos, Notoginseng Flos had 2 and 7 information sites, respectively. The neighbor join system cluster tree showed that the species were clustered independently into one branch, with Ginseng Flos, and Quinquefolii Flos clustered as a large branch and juxtaposed with Notoginseng Flos. Ginseng Flos was the same as Quinquefolii Flos secondary structure, but with Notoginseng Flos there were three different positions but there were A, B and C differences between arm Ⅳ and arm Ⅰ of Notoginseng Flos. CONCLUSION The high-throughput sequencing technology can accurately identify Ginseng Flos, Quinquefolii Flos and Notoginseng Flos, and has a strong ability to identify adulterated samples, which provides a certain idea for the identification of commercial Ginseng Flos.