Greg Budoff1, Loka Thangamathesvaran1, Marco A Zarbin1, Neelakshi Bhagat2. 1. Institute of Ophthalmology and Visual Science, Rutgers - New Jersey Medical School, Newark, New Jersey. 2. Institute of Ophthalmology and Visual Science, Rutgers - New Jersey Medical School, Newark, New Jersey. Electronic address: bhagatne@njms.rutgers.edu.
Abstract
PURPOSE: Population-based reports regarding bacterial endogenous endophthalmitis (BEE) are scarce in the literature. Our study sought to investigate the incidence of BEE, relative risk of comorbid conditions, and treatment patterns using a nationally representative database, the National Inpatient Sample (NIS). DESIGN: Retrospective cross-sectional study. PARTICIPANTS: Inpatients with bacteremia with and without BEE in the NIS database between 2002 and 2013. METHODS: Demographic and comorbid medical conditions were evaluated in bacteremic patients who demonstrated BEE and in those who did not; furthermore, logistic regression was performed to predict the relative risk of comorbid conditions in BEE developing. Chi-square analysis was used to determine if certain organisms were more likely to cause BEE and their role in treatment method. Statistical analysis was performed using SPSS software version 22 (IBM, Armonk, NY). MAIN OUTCOME MEASURES: Comorbidities, associated microorganisms, and surgical management. RESULTS: Among 3 907 204 bacteremic inpatients, 1793 (0.05%) were found to have BEE. Patients were more likely to demonstrate BEE in the presence of endocarditis (P < 0.001), bacterial meningitis (P < 0.001), pyogenic arthritis (P < 0.001), internal organ abscess (P < 0.001), osteomyelitis (P < 0.001), prostatitis (P < 0.001), diabetic retinopathy (P < 0.001), and vascular catheter placement or infection (P < 0.001). Diabetes mellitus without complications was not associated with an increased risk of BEE. Methicillin-resistant Staphylococcus aureus (MRSA), streptococcal, and pneumococcal bacteremia were found to significantly increase the risk of BEE (P < 0.001 for all), and the incidences of BEE associated with these organisms were 0.19%, 0.09%, and 0.07%, respectively. Fewer than 10% of patients with BEE underwent vitrectomy, and 1.6% of BEE patients underwent enucleation. CONCLUSIONS: Bacterial endogenous endophthalmitis was observed in approximately 1 in 2000 inpatients with bacteremia in this large cohort of NIS inpatients. It was seen more commonly with MRSA, streptococcal, and pneumococcal species. Endocarditis, meningitis, prostatitis, osteomyelitis, infected vascular catheter, internal organ abscess, and diabetic retinopathy showed a significant risk (P < 0.001) of being associated with BEE.
PURPOSE: Population-based reports regarding bacterial endogenous endophthalmitis (BEE) are scarce in the literature. Our study sought to investigate the incidence of BEE, relative risk of comorbid conditions, and treatment patterns using a nationally representative database, the National Inpatient Sample (NIS). DESIGN: Retrospective cross-sectional study. PARTICIPANTS: Inpatients with bacteremia with and without BEE in the NIS database between 2002 and 2013. METHODS: Demographic and comorbid medical conditions were evaluated in bacteremic patients who demonstrated BEE and in those who did not; furthermore, logistic regression was performed to predict the relative risk of comorbid conditions in BEE developing. Chi-square analysis was used to determine if certain organisms were more likely to cause BEE and their role in treatment method. Statistical analysis was performed using SPSS software version 22 (IBM, Armonk, NY). MAIN OUTCOME MEASURES: Comorbidities, associated microorganisms, and surgical management. RESULTS: Among 3 907 204 bacteremic inpatients, 1793 (0.05%) were found to have BEE. Patients were more likely to demonstrate BEE in the presence of endocarditis (P < 0.001), bacterial meningitis (P < 0.001), pyogenic arthritis (P < 0.001), internal organ abscess (P < 0.001), osteomyelitis (P < 0.001), prostatitis (P < 0.001), diabetic retinopathy (P < 0.001), and vascular catheter placement or infection (P < 0.001). Diabetes mellitus without complications was not associated with an increased risk of BEE. Methicillin-resistant Staphylococcus aureus (MRSA), streptococcal, and pneumococcal bacteremia were found to significantly increase the risk of BEE (P < 0.001 for all), and the incidences of BEE associated with these organisms were 0.19%, 0.09%, and 0.07%, respectively. Fewer than 10% of patients with BEE underwent vitrectomy, and 1.6% of BEE patients underwent enucleation. CONCLUSIONS: Bacterial endogenous endophthalmitis was observed in approximately 1 in 2000 inpatients with bacteremia in this large cohort of NIS inpatients. It was seen more commonly with MRSA, streptococcal, and pneumococcal species. Endocarditis, meningitis, prostatitis, osteomyelitis, infected vascular catheter, internal organ abscess, and diabetic retinopathy showed a significant risk (P < 0.001) of being associated with BEE.