Objective: To compare clinical outcomes of patients treated by female surgeons versus those treated by male surgeons. Background: It remains unclear as to whether surgical performance and outcomes differ between female and male surgeons. Methods: We conducted a meta-analysis to compare patients’ clinical outcomes—including patients’ postoperative mortality, readmission, and complication rates—between female versus male surgeons. MEDLINE, Embase, CENTRAL, ICTRP, and ClinicalTrials.gov were searched from inception to September 8, 2022. The update search was conducted on July 19, 2023. We used random-effects models to synthesize data and GRADE to evaluate the certainty. Results: A total of 15 retrospective cohort studies provided data on 5,448,121 participants. We found that patients treated by female surgeons experienced a lower postoperative mortality compared with patients treated by male surgeons [8 studies; adjusted odds ratio (aOR), 0.93; 95% CI, 0.88–0.97; I ²=27%; moderate certainty of the evidence]. We found a similar pattern for both elective and nonelective (emergent or urgent) surgeries, although the difference was larger for elective surgeries (test for subgroup difference P=0.003). We found no evidence that female and male surgeons differed for patient readmission (3 studies; aOR, 1.20; 95% CI, 0.83–1.74; I ²=92%; very low certainty of the evidence) or complication rates (8 studies; aOR, 0.94; 95% CI, 0.88–1.01; I ²=38%; very low certainty of the evidence). Conclusion: This systematic review and meta-analysis suggests that patients treated by female surgeons have a lower mortality compared with those treated by male surgeons.

Background: Current guidelines for community-acquired pneumonia (CAP) include ampicillin-sulbactam as an initial treatment option, though they do not mandate routine coverage of anaerobic organisms. This study aimed to compare the effectiveness of ampicillin-sulbactam with that of ceftriaxone as initial treatment for CAP in older adults. Methods: This study was conducted using the target trial emulation framework, using a nationwide Japanese database (May 2010–June 2023). The study included patients aged ≥65 years, admitted to Diagnosis Procedure Combination hospitals for CAP, who received either ampicillin-sulbactam or ceftriaxone as the initial treatment. The exposure group received ampicillin-sulbactam, while the control group received ceftriaxone, both on the day of hospitalization. The primary outcome was in-hospital mortality; the secondary outcome was the development of Clostridioides difficile infection during hospitalization. Results: The study included 26 633 older patients hospitalized with CAP, with 14 906 receiving ampicillin-sulbactam and 11 727 receiving ceftriaxone as initial treatment. After inverse probability of treatment weighting, the ampicillin-sulbactam group was associated with a higher in-hospital mortality rate than the ceftriaxone group (10.5% vs 9.0%, respectively; adjusted risk difference, 1.5% [95% confidence interval, .7%–2.4%]; adjusted odds ratio, 1.19 [1.08–1.31]). The incidence of C difficile infection was numerically higher in the ampicillin-sulbactam group (0.6% vs 0.4%; adjusted risk difference, 0.2% [95% confidence interval, .0%–.4%]; adjusted odds ratio, 1.45 [.99–2.11]). These results were consistent among patients with risk factors for aspiration. Conclusions: In older patients with CAP, initial treatment with ampicillin-sulbactam was associated with higher mortality compared to treatment with ceftriaxone.

Systematic reviews (SRs) are a critical component of evidence-based medicine, but the process of screening titles and abstracts is time-consuming. This study aimed to develop and externally validate a method using large language models to classify abstracts for diagnostic test accuracy (DTA) systematic reviews, thereby reducing the human workload. We used a previously collected dataset for developing DTA abstract classifiers and applied prompt engineering. We developed an optimized meta-prompt for Generative Pre-trained Transformer (GPT)-3.5-turbo and GPT-4 to classify abstracts. In the external validation dataset 1, the prompt with GPT-3.5 turbo showed a sensitivity of 0.988, and a specificity of 0.298. GPT-4 showed a sensitivity of 0.982, and a specificity of 0.677. In the external validation dataset 2, GPT-3.5 turbo showed a sensitivity of 0.919, and a specificity of 0.434. GPT-4 showed a sensitivity of 0.806, and a specificity of 0.740. If we included eligible studies from among the references of the identified studies, GPT-3.5 turbo had no critical misses, while GPT-4 had some misses. Our study indicates that GPT-3.5 turbo can be effectively used to classify abstracts for DTA systematic reviews. Further studies using other dataset are warranted to confirm our results. Additionally, we encourage the use of our framework and publicly available dataset for further exploration of more effective classifiers using other LLMs and prompts (https://github.com/youkiti/ARE/). What is already known - Title and abstract screening in systematic reviews (SRs) consumes significant time. - Several attempts using machine learning to reduce this process in diagnostic test accuracy (DTA) SRs exist, but they have not yielded positive results in external validation. What is new - We aimed to develop and externally validate optimized meta-prompt for GPT-3.5-turbo and GPT-4 to classify abstracts for DTA SRs. - Through an iterative approach across three training datasets, an optimal meta-prompt capable of identifying DTA studies with remarkable sensitivity and specificity was developed. - The accuracy reproduced in the external validation datasets. Potential Impact for Readers - The developed meta-prompt can lessen the need for humans to read abstracts for DTA SRs, saving significant time and resources.