目的 评估多部门合作下，不同抗生素使用方案在关节骨科中的应用，对优化护理人力资源配置、降低医疗耗材支出方面的效果，并验证其预防感染的安全性是否不劣于传统方案。 方法 采用回顾性队列研究方法，选取2021年5月10日至2024年5月22日广东省中医院膝关节病科的患者1 948例，对其中303例（髋关节44例，膝关节259例）的炎症因子进行亚组分析，按抗生素使用方案分为单次组（972例）和多次组（976例），单次组术后仅单次给药，多次组术后多次给药。主要观察指标：术后12个月假体周围感染（PJI）发生率。次要观察指标：（1）炎症因子：白细胞计数（white blood cell count， WBC）、中性粒细胞计数、C反应蛋白（C-reactive protein， CRP）、血沉（ESR）；（2）成本效益指标：抗生素输液费用（单项住院费用）；（3）护理负荷：输液距离加拔针距离。统计分析采用logistic回归、Mann-Whitney U检验等。 结果 单次组与多次组PJI分别为1.5%（15/972）和1.3%（13/976），差异无统计学意义（OR = 1.039， 95%CI： 0.927 ~ 1.164， P = 0.512）。亚组分析结果：髋置换术使用头孢呋辛钠炎症指标（WBC、中性粒细胞、CRP、ESR），在各时点差异均有统计学意义（P < 0.05）；组间差异均无统计学意义（P > 0.05）；膝关节置换术使用头孢唑啉钠的患者（WBC、中性粒细胞、CRP、ESR），炎症指标在各时点差异均有统计学意义（P < 0.05）；组间差异CRP两组具有统计学意义（P < 0.05），其余指标组间差异均无统计学意义（P > 0.05）。膝关节置换术使用头孢呋辛钠（WBC、中性粒细胞、CRP、ESR），在各时点差异均有统计学意义（P < 0.05）；ESR组间差异不具有可比性（P < 0.05），其余指标组间差异均无统计学意义（P > 0.05）。成本效益指标：单次组抗生素费用中位数显著低于多次组（74.13元vs.164.16元，Z = -39.090，P < 0.001），护理负荷（输液 + 拔针距离）中位数减少67.1%（298.4 m vs. 907.8 m， Z = -37.536，P < 0.001）。 结论 单次抗生素方案在感染预防效果与多次方案相当，且显著降低医疗成本（费用减少55%）及护理负荷（工作量减少67%），建议临床推广，本研究为单中心回顾性研究，未来需多中心前瞻性研究进一步验证。

Objective To evaluate the effects of different antibiotic regimens applied in orthopedic joint surgery through multi-departmental collaboration on optimizing nursing human resource allocation and reducing medical consumables expenditure， and to verify whether its safety in preventing infection is non-inferior to the traditional regimen. Methods A retrospective cohort study was conducted. A total of 1 948 patients from the Department of Knee Surgery between May 10， 2021， and May 22， 2024， were selected. A subgroup analysis of inflammatory markers was performed on 303 of these patients （44 hip joints， 259 knee joints）. Patients were divided into a single-dose group （972 cases） and a multiple-dose group （976 cases） based on the antibiotic regimen. The single-dose group received only one postoperative dose of antibiotics， while the multiple-dose group received multiple postoperative doses. The primary outcome was the relationship between the number of antibiotic packages used and the incidence of prosthetic joint infection （PJI） within 12 months postoperatively. Secondary outcomes included inflammatory markers： white blood cell count （WBC）， neutrophil count， C-reactive protein （CRP）， and erythrocyte sedimentation rate （ESR）； cost-effectiveness indicators： antibiotic infusion cost （as a single hospitalization expense）； and nursing workload： infusion distance plus needle removal distance. Statistical analyses included logistic regression and the Mann-Whitney U test. Results The overall incidence of prosthetic joint infection （PJI） at 12 months postoperatively was low. The PJI rates in the single-dose and multiple-dose groups were 1.5% （15/972） and 1.3% （13/976）， respectively， with no statistically significant difference （OR = 1.039， 95%CI： 0.927 ~ 1.164， P = 0.512）. Subgroup analysis results： For hip arthroplasty patients using cefuroxime sodium， inflammatory markers （WBC， neutrophils， CRP， ESR） showed statistically significant differences at various time points within each group （P < 0.05）， but no statistically significant intergroup differences were observed （P > 0.05）. For knee arthroplasty patients using cefazolin sodium， inflammatory markers （WBC， neutrophils， CRP， ESR） showed statistically significant differences at various time points within each group （P < 0.05）. The intergroup difference for CRP was not comparable （P < 0.05）， while no statistically significant intergroup differences were found for the remaining markers （P > 0.05）. For knee arthroplasty patients using cefuroxime sodium， inflammatory markers （WBC， neutrophils， CRP， ESR） showed statistically significant differences at various time points within each group （P < 0.05）. The intergroup difference for ESR was not comparable （P < 0.05）， while no statistically significant intergroup differences were observed for the remaining markers （P > 0.05）. Cost-effectiveness analysis： The median antibiotic cost in the single-dose group was significantly lower than that in the multiple-dose group （74.13 RMB vs. 164.16 RMB， Z = -39.090， P < 0.001）. The nursing workload （infusion distance + needle removal distance） was reduced by a median of 67.1% （298.4 meters vs. 907.8 meters， Z = -37.536， P < 0.001）. Conclusions The single-dose antibiotic regimen is comparable to the multiple-dose regimen in preventing infections， while significantly reducing medical costs （by 55%） and nursing workload （by 67%）. Its clinical adoption is recommended. This study is a single-center retrospective investigation， and future multi-center prospective studies are warranted for further validation.