Objective:To evaluate the performance of three mainstream large language models (LLMs) in the review of drug reimbursement indications in hospitals, and to explore their potential in improving audit quality and efficiency, thereby safeguarding the medical insurance fund.Methods:A total of 3 247 outpatient prescription records were retrospectively collected from a specialized oncology hospital between January 2, 2022, and June 30, 2023. Manual assessment of the consistency between clinical diagnoses and drug reimbursement indications was used as the gold standard. Three LLMs, Baidu′s ERNIE Bot, Alibaba′s Tongyi Qianwen, and OpenAI′s ChatGPT-4o, were evaluated on the same task. Performance metrics included accuracy, precision, sensitivity, specificity, F1 score, and area under the curve (AUC).Results:The ERNIE Bot model returned 3 242 valid data, which took 314 min; The Tongyi Qianwen model returned a total of 3 162 valid data, taking 384 min; The ChatGPT-4o model returned a total of 3 218 valid data, taking 150 min. ChatGPT-4o demonstrated the best performance, with an accuracy of 88.41%, precision of 60.48%, sensitivity of 78.75%, specificity of 90.24%, F1 score of 0.68, and an AUC of 0.88.Conclusions:LLMs demonstrate stable performance in determining whether prescriptions align with reimbursement indications, with ChatGPT-4o approaching human-level accuracy and exhibiting more conservative specificity. These findings suggest that LLMs have practical value as auxiliary tools in drug indication reviews, contributing to improved audit efficiency and more refined management of medical insurance funds.

Objective:To evaluate the performance of three mainstream large language models (LLMs) in the review of drug reimbursement indications in hospitals, and to explore their potential in improving audit quality and efficiency, thereby safeguarding the medical insurance fund.Methods:A total of 3 247 outpatient prescription records were retrospectively collected from a specialized oncology hospital between January 2, 2022, and June 30, 2023. Manual assessment of the consistency between clinical diagnoses and drug reimbursement indications was used as the gold standard. Three LLMs, Baidu′s ERNIE Bot, Alibaba′s Tongyi Qianwen, and OpenAI′s ChatGPT-4o, were evaluated on the same task. Performance metrics included accuracy, precision, sensitivity, specificity, F1 score, and area under the curve (AUC).Results:The ERNIE Bot model returned 3 242 valid data, which took 314 min; The Tongyi Qianwen model returned a total of 3 162 valid data, taking 384 min; The ChatGPT-4o model returned a total of 3 218 valid data, taking 150 min. ChatGPT-4o demonstrated the best performance, with an accuracy of 88.41%, precision of 60.48%, sensitivity of 78.75%, specificity of 90.24%, F1 score of 0.68, and an AUC of 0.88.Conclusions:LLMs demonstrate stable performance in determining whether prescriptions align with reimbursement indications, with ChatGPT-4o approaching human-level accuracy and exhibiting more conservative specificity. These findings suggest that LLMs have practical value as auxiliary tools in drug indication reviews, contributing to improved audit efficiency and more refined management of medical insurance funds.

Objective:To explore the clinical value of autoantibodies in patients with liver disease.Methods:We retrospectively analyzed the data of 1 495 outpatients or inpatients with liver disease in Beijing Ditan Hospital of Capital Medical University from August 2020 to April 2021. Indirect immunofluorescence and Western blot were used to detect antinuclear antibody (ANA) and antinuclear antibodies (ANAs).Results:ANA and ANAs were positive in patients with liver diseases of various etiologies. Among 1 495 patients with liver disease, 494 cases were ANA positive, the positive rate was 33.04%; 573 cases were positive for ANAs, the positive rate was 38.33%. The positive rate of ANA in the immune liver disease group (63.37%) was higher than that in the viral, alcoholic, fatty liver, confounding factors and other liver disease groups, and the difference was statistically significant ( P<0.01). The ANA positive rate between the viral, alcoholic, fatty liver, and confounding factor groups was statistically significant ( χ2=19.823, P<0.01), the positive rate of ANAs in the immune liver disease group (80.23%) was higher than that in other liver disease groups, and the difference was statistically significant ( P<0.05). The antibody titer of immune liver disease group was mainly 1∶1000, and other liver disease etiology groups was mainly 1∶100. The two most common fluorescent karyotypes in liver disease groups of different etiologies are cytoplasmic and nuclear granular types. The most common specific antibody in the immune liver disease group was anti-mitochondrial antibody type 2 (anti-AMA-M2) antibody, the most common anti-Ro-52 antibody in viral, drug-induced, complex etiology, and other etiological groups, and the most common anti-SSA antibody in alcoholic liver disease. Anti-SSA antibody (17.44%), anti-SSB antibody (9.30%), anti-CENP-B antibody (22.09%), anti-Ro-52 antibody (41.28%), anti-AMA-M2 antibody (51.74%) were positive in immune liver disease group, The rate was higher than that of other liver disease etiology groups, and the difference was statistically significant ( P<0.01). When the ANA fluorescence karyotype is nuclear granule type, the positive rate of anti-CENP-B antibody, anti-Ro-52 antibody, and anti-AMA-M2 antibody in the immune liver disease group was higher than that in the viral liver disease group ( P<0.01), The positive rate of anti-Ro-52 antibody was higher than that of drug-induced liver disease group ( P<0.05). Conclusions:The ANA titer of autoimmune liver disease was mainly (1∶1 000). ANAs were mainly positive for anti-SSA antibody, anti-SSB antibody, anti-CENP-B antibody, anti-Ro-52 antibody, and anti-AMA-M2 antibody, especially anti-AMA-M2 antibody. When combined with ANA fluorescent karyotype and ANAs for analysis, if the fluorescent karyotype is nuclear particle type, the positive anti-Ro-52 antibody in ANAs is more valuable in distinguishing immunity from viral and drug-induced liver diseases.

Objective To construct and identify a recombinant Bifidobacteria(Bb) (pGEX-TSO45W-4B) vaccine of Taenia solium.Methods The TSO45W-4B gene of Taenia solium was synthesized.It was expected that the fragment length of the gene was 351 bp.The gene was cloned into Escherichia coli-Bifidobacteria(Bb) shutde vector pGEX-1λT to construct a recombinant plasmid pGEX-TSO45W-4B.The recombinant plasmid was electroporated into Bb to construct a recombinant Bb (pGEX-TSO45W-4B) vaccine.The vaccine was identified by restriction analysis,PCR and sequencing,which would demonstrate that the constructed recombinant Bb (pGEX-TSO45W-4B) vaccine of Taenia solium contained the gene of fragment length of 351 bp.Results The TSO45W-4B gene of 351 bp was synthesized.Restriction analysis,PCR and sequencing results showed that the recombinant Bb (pGEX-TSO45W-4B) vaccine of Taenia solium was successfully constructed and contained the gene fragment of 351 bp.Conclusion The recombinant Bb (pGEX-TSO45W-4B) vaccine of Taenia solium is successfully constructed,which lays a foundation for the study of expression and immunogenicity of the vaccine.