We read with great interest the article by Vaziri and associates describing risk factors predictive of endogenous endophthalmitis among hospitalized patients with hematogenous infections in the United States. In their study, the leading predictors were infectious meningitis, endocarditis, and abscesses; comorbidities indicative of immunocompromise (including HIV/AIDS, lymphoma/leukemia, and diabetes with systemic complications); intensive care unit admission; and longer hospital stays. Although the study was concise, we have some concerns about how the study was presented.

First, the length of hospital stay generally depends on the primary infectious disease. Patients with meningitis, endocarditis, and internal organ abscesses require longer intravenous antibiotic treatment, which is the main treatment for eradicating pathogens. Sepsis with disseminated intravascular coagulation is a more severe type of bacteremia/fungemia. Patients with sepsis have a higher risk of developing endogenous endophthalmitis, and they require regular or neonatal intensive care unit admission and longer hospitalization. We would like to know whether patients with sepsis have a higher odds ratio.

Second, urinary tract infection (UTI) was associated with a statistically significant (but not clinically significant) reduction in the odds of endophthalmitis among the patients with bacteremia/fungemia in that study. UTIs include urethritis, cystitis, pyelonephritis, and renal abscesses. Patients with bacteremia/fungemia typically have positive urine culture results, and UTI may be diagnosed confoundingly. If other infectious sources are not identified, UTI would be diagnosed as the primary infectious source. As the authors stated, pyelonephritis and renal abscess are more likely to be associated with hematogenous spread, particularly the bacteria Klebsiella pneumoniae and Escherichia coli . Patients with pyelonephritis and renal abscess also require longer hospital stays for intravenous antibiotic treatment.

Third, the 2 patients with diabetes and ophthalmic manifestations, as well as the 19 patients with diabetes, developed endogenous endophthalmitis. We would like to know whether the condition of the 17 patients with chronic kidney disease/renal failure was related to diabetes or other kidney disease. Because patients with diabetes develop chronic kidney disease/renal failure, they typically present ophthalmic manifestations. Furthermore, the research method in the study considered only first hospitalization data, and ophthalmic manifestations might not be listed on the final diagnosis. The effect of diabetes with ophthalmic manifestation on endogenous endophthalmitis may have been underestimated or overestimated. Additionally, poor control of diabetes (HbA1c >9%) is typically associated with a higher risk of endogenous endophthalmitis in diabetic patients with bacteremia/fungemia.

Finally, we are also interested in the primary risk factors in the various age groups. Different age groups, particularly the pediatric group, have distinct risk factors. People younger than 17 years generally have no diabetes with ophthalmic manifestations. Prematurity with low birth weight, pyogenic pneumonia, and leukemia could be critical risk factors. Endogenous endophthalmitis is an uncommon but potentially sight-threatening disease, and patients with multiple risk factors should be consulted early. If the authors could provide further information, it would assist internists and ophthalmologists in screening the endogenous endophthalmitis.