ObjectiveTo observe the clinical features and treatment outcomes of patients with intraocular foreign bodies with endophthalmitis, and analyze the prognostic factors affecting the anatomic and visual outcomes of patients. MethodsA retrospective clinical study. A total of 1 704 patients (1 704 eyes) with intraocular foreign body at Eye Hospital, Wenzhou Medical University from January 2015 to June 2024 were included in this study. Endophthalmitis was diagnosed in 263 eyes (15.4%, 263/1 704). Patients who lost follow-up in our hospital after surgery were excluded, 155 patients with 155 eyes were finally included in the study. Uncorrected visual acuity (UCVA) examination was performed before operation. Best corrected visual acuity (BCVA) examination was performed both after the first stage debridement and during follow-up. The visual acuity test is performed using a standard logarithmic visual acuity chart, which is statistically converted to logarithm of the minimum angle of resolution (logMAR) visual acuity. Demographic characteristics (gender, age), trauma characteristics (time of injury, occupation characteristics, nature of foreign body), anatomical injury (wound zoning, nature of infection, etc.), clinical treatment (interval between operation and injury, rate of second operation, etc.) and outcome (vision outcome, complications, anatomic outcome, etc.) were recorded. Prophylactic intravitreous injection of 10 mg/ml of cefazolin sodium 0.1ml (including 1 mg of cefazolin sodium) was given on the basis of perioperative systemic administration of cefazolin sodium from 2022. Anatomical outcomes included anatomical reduction, silicone oil-dependent, and ophthalmectomy. The visual outcomes of the patients were categorized into three groups based on the best-corrected visual acuity at the final follow-up: visual acuity worse than 0.05, visual acuity between 0.05 and 0.3, and visual acuity better than 0.3. Generalized linear mixed model (GLMM) was used to analyze the correlation between the timing of treatment, nature of foreign body, nature of infection, number of operations, location of injury and the anatomic and visual outcomes of patients. ResultsOf 155 patients, 149 were males and 6 were females, mean age was (45.7±12.9) years, patients with monocular injury. Magnetic, non-magnetic, unidentified metal and vegetable, mineral, animal and unidentified foreign bodies were 102 (65.8%, 102/155), 2 (1.3%, 2/155), 28 (18.1%, 28/155), 1 (0.6%, 1/155), 12 (7.7%, 12/155), 7 (4.5%, 7/155), 3 (1.9%, 3/155) cases, respectively. The time between injury and removal of foreign body was (98.1±359.5) h. The foreign bodies were removed in 136 eyes (87.2%, 136/155) in the primary surgery, 67 cases combined with debridement and suture, 68 cases combined with pars plana vitrectomy (PPV), and 1 case suffered ophthalmectomy. The slide and culture results revealed that the eyes positive for bacteria and those positive for a mixed infection of bacteria and fungi were 80 (51.2%, 80/155) and 2 (1.3%, 2/155) eyes, respectively; 73 eyes (46.8%, 73/155) were negative. Among the 80 eyes positive for bacteria, staphylococcus epidermidis and bacillus cereus were found in 26 (32.5%, 26/80) and 23 (28.8%, 23/80) eyes, respectively. Drug sensitivity testing indicated that vancomycin, gentamicin and amikacin had low drug resistance (1.79%, 6.67%, 0.0%, respectively). The mean preoperative logMAR UCVA was 1.67±0.79. In the outcome of visual function, 78, 26 and 51 patients with visual acuity ＜0.05, 0.05-0.3, ＞0.3, respectively. At the last follow-up, there were 56 cases (36.1%, 56/155) of silicone oil dependence, 93 cases (60%, 93/155) of anatomic reduction, and 6 cases (3.9%, 6/155) suffered ophthalmectomy, the mean intraocular pressure was (13.6±6.1) mm Hg (1 mm Hg= 0.133 kPa). Preoperative visual acuity was strongly correlated with visual outcomes (F=6.896, P=0.001). Preoperative visual acuity (F=5.310, P=0.023) and surgical method (F=20.448, P＜0.001) were closely related to the anatomical outcome, while age, treatment time, foreign body nature, wound zoning, infection nature, and foreign body removal time had no statistical correlation with the anatomic and functional outcome (P＞0.05). During 2015 to 2024, the incidence of intraocular foreign body-related endophthalmitis was 12.5%-22.7%, which showed a fluctuating upward trend. The incidence of endophthalmitis increased during 2022 to 2024 compared with the period from 2019 to 2021, but no statistically difference was found (χ2=3.856, P=0.05). ConclusionsThe incidence of intraocular foreign body related endophthalmitis was 15.4%. Staphylococcus epidermidis and Bacillus cereus are the first and second pathogenic bacteria. The incidence of endophthalmitis is not significantly reduced with intravitreal injection of cefazolin sodium. Preoperative UCVA and surgical method were closely related to the anatomic outcome of patients.