Abstract
Purpose:
This work describes the characteristics, treatment, and outcomes of endogenous endophthalmitis (EE) cases over 13 years at a tertiary care hospital.
Methods:
This retrospective case series included EE cases from 2006 to 2018. Eyes were categorized by cultured organism and initial treatment status.
Results:
Forty-five eyes of 38 patients were identified with EE. Mean age was 48 ± 12 years and approximately three-fourths were male. Mean visual acuity (VA) at presentation was logMAR 1.86 ± 0.83. More than 50% of patients had a history of intravenous drug use. In a quarter of patients, no systemic source of infection was identified; 90% of these patients had a history of intravenous drug use. Bilateral infections occurred in 18% of participants. Of positive cultures, 29%, 56%, and 13% were fungal, gram-positive, and gram-negative in etiology, respectively. Eyes with fungal infections had a better presenting VA and were more likely to have better than 20/100 final VA vs those with bacterial infections. All patients received IV antibiotics; additionally 29% of eyes underwent initial combined pars plana vitrectomy with intravitreal antibiotic injection while 51% of eyes initially received bedside intravitreal antibiotic injection. There was no significant difference in final VA with respect to initial mode of treatment.
Conclusions:
Initial treatment of intravitreal antibiotics with or without pars plana vitrectomy did not show any difference in the average final VA in eyes with EE. Eyes with fungal infections had more favorable visual outcomes than those with bacterial infections. Almost 1 in 5 EE eyes were enucleated.
Keywords: endophthalmitis, retina, bacteremia, fungemia, septicemia
Introduction
Endogenous endophthalmitis (EE), also known as metastatic endophthalmitis, is an uncommon condition in which a blood-borne infection seeds the eye. Predisposing conditions include immunosuppression, diabetes, endocarditis, and intravenous drug use (IVDU), among others. 1 -15 The causative agent varies worldwide, with gram-positive and fungal etiologies more common in North America and Europe 1,5 -8,10 -14 and gram-negative species more typical in Asia. 2,9,15 Population-based reports suggest an incidence of 0.04% to 0.4% for septicemic patients. 13,14 Given the rarity of EE, an optimal treatment regimen has not been established.
In this study we sought to characterize our experience with EE by evaluating the presentation, microbiologic causes, treatment course, and outcomes in a tertiary care urban university hospital practice setting.
Methods
This retrospective medical record review study included all patients with presumed EE treated at University Hospital in Newark, New Jersey, between January 1, 2006, and December 31, 2018. Patients were excluded if there was a history of recent ocular trauma, recent ocular surgery, or a primary ocular focus of infection.
Records were reviewed for risk factors, comorbidities, examination findings, microbial cultures, treatments, and visual outcomes. Bacterial or fungal cultures were considered positive if microbial growth was documented in either ocular or systemic (blood, urine, or tissue) cultures. Patients without positive cultures were presumed to have fungal endophthalmitis if the clinical presentation was strongly suggestive of fungal disease. Otherwise, culture-negative patients were labeled unknown and excluded from statistical analysis.
All mean values are reported with ± SD. Visual acuities (VAs), recorded in the medical records using Snellen charts, were converted to logMAR for statistical calculations. 16 LogMAR values of 2.0, 2.3, 2.6, and 2.9 were used for VAs of counting fingers (CF), hand motions (HM), light perception (LP), and no light perception (NLP), respectively. 17 Analyses of variance and t tests were performed when comparing mean values across 2 or 3 groups, respectively. Fisher exact test was used to compare categorical values. The type 1 error rate was set at .05.
Results
Forty-five eyes of 38 patients with EE were identified and bilateral infections were seen in 7 (18%) individuals. The mean age was 48 ± 12 years (range, 24-70 years). Twenty-seven patients (71%) were male. The right eye was involved in 23 eyes (51%). Mean duration of symptoms prior to presentation was 12 ± 25 days (median, 6 days).
The microbial etiology was fungal in 13 (29%) eyes of 11 patients, gram-positive in 25 (56%) eyes of 22 patients, gram-negative in 6 (13%) eyes of 4 patients, and unknown in 1 (2%) eye of 1 patient. Three (23%) of the 13 eyes diagnosed as having fungal endophthalmitis had negative culture findings, and none of these culture-negative eyes were in patients with bilateral disease. One patient had negative systemic and ocular culture results, and the etiology of this infection was unknown.
Comorbidities and presumed sources are shown in Table 1. More than 50% of all patients had a history of IVDU. The presumed source was unknown in 26% of patients, although 90% of these patients had a history of IVDU.
Table 1.
Comorbidities and Presumed Source of Infection.
| Comorbidity | No. of patients | % |
|---|---|---|
| IVDU | 20 | 53 |
| Diabetes mellitus | 13 | 34 |
| Hepatitis C | 8 | 21 |
| Cancer | 6 | 16 |
| Hemodialysis | 5 | 13 |
| HIV | 4 | 10 |
| Cirrhosis | 3 | 8 |
| Presumed source | No. of patients | % |
| Endocarditis | 7 | 18 |
| Subcutaneous infection | 7 | 18 |
| Catheter site infection | 3 | 8 |
| Osteomyelitis | 3 | 8 |
| Meningitis | 2 | 5 |
| Urosepsis | 2 | 5 |
| Prostatitis | 1 | 3 |
| Pneumonia | 1 | 3 |
| Colonic perforation | 1 | 3 |
| Peritonitis | 1 | 3 |
| Unknown | 10 | 26 |
Abbreviation: IVDU, intravenous drug use.
Presentation
Presenting symptoms included decreased vision (92% of patients), ocular pain (32%), redness (18%), floaters (13%), and photophobia (5%). Historical and examination findings are shown in Table 2. Ocular cultures were performed either via bedside vitreous tap or in the operating room via pars plana vitrectomy (PPV) and yielded positive results in 16 (46%) of 35 eyes; the remaining eyes had either a dry vitreous tap or no vitreous culture performed. Findings from systemic cultures (blood, urine, or tissue) were significantly more likely to be negative in patients with fungal etiologies when compared with gram-positive cases (P = .002), but there was no significant difference when compared with gram-negative (P = .28) etiologies. Table 3 lists all cultured organisms.
Table 2.
Historical and Examination Findings by Etiology.
| Fungal | Gram-positive | Gram-negative | ||||||
|---|---|---|---|---|---|---|---|---|
| Characteristic | ± | na | % | na | % | na | % | P, Fisher exact test |
| IVDU, patients | + | 8 | 73 | 10 | 45 | 1 | 25 | .23 |
| – | 3 | 27 | 12 | 55 | 3 | 75 | ||
| Hypopyon | + | 3 | 23 | 18 | 72 | 1 | 17 | .003 |
| – | 10 | 77 | 7 | 28 | 5 | 83 | ||
| View of fundus | + | 12 | 92 | 7 | 28 | 1 | 17 | < .001 |
| – | 1 | 8 | 18 | 72 | 5 | 83 | ||
| Ocular cultureb | + | 8 | 67 | 5 | 26 | 3 | 75 | .04 |
| – | 4 | 33 | 14 | 74 | 1 | 25 | ||
| Systemic culture, patients | + | 4 | 36 | 20 | 91 | 3 | 75 | .03 |
| – | 7 | 64 | 2 | 9 | 1 | 25 | ||
| Known sepsis at examination, patients | + | 4 | 36 | 17 | 77 | 2 | 50 | .07 |
| – | 7 | 64 | 5 | 23 | 2 | 50 | ||
Abbreviation: IVDU, intravenous drug use.
a Numbers represent eyes unless otherwise specified in the left column.
b Eyes with dry taps were excluded.
Table 3.
Cultured Organisms Across All Culture Sources.
| Organism | No. | % | Bilateral infection, No. |
|---|---|---|---|
| MRSA | 10 | 22 | 1 |
| Candida albicans | 7 | 16 | 2 |
| Klebsiella pneumoniae | 5 | 11 | 2 |
| MSSA | 4 | 9 | 1 |
| Streptococcus agalactiae | 2 | 4 | 0 |
| Viridins streptococci | 2 | 4 | 0 |
| Staphylococcus auricularis | 2 | 4 | 1 |
| Streptococcus pneumoniae | 2 | 4 | 0 |
| Aspergillus fumigatus | 1 | 2 | 0 |
| Group G streptococci | 1 | 2 | 0 |
| Streptococcus pyogenes | 1 | 2 | 0 |
| Serratia marcescens | 1 | 2 | 0 |
| Bacillus cereus | 1 | 2 | 0 |
| Culture negative | 6 | 13 | 0 |
Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive S aureus.
The mean VA at presentation across all eyes was logMAR 1.86 ± 0.83 (Snellen equivalent, 20/1500). Eyes with fungal infection were found to present with significantly better mean VA than those with bacterial etiologies (P < .001 and P = .03 when comparing eyes with fungal etiologies to those with gram-positive and gram-negative etiologies, respectively). Table 4 presents VA data stratified by etiology.
Table 4.
Visual Acuity by Type of Microorganism.
| Mean VA in logMAR | Fungal | Gram-positive | Gram-negative | P | Test | |||
|---|---|---|---|---|---|---|---|---|
| Initial | 1.21 ± 0.77 | 2.11 ± 0.66 | 2.22 ± 0.98 | .002 | ANOVA | |||
| Final | 1.28 ± 1.10 | 2.00 ± 0.77 | 2.17 ± 1.20 | .062 | ANOVA | |||
| P, t test | .85 | .36 | .83 | – | – | |||
| Mean change | –0.07 ± 1.05 | –0.11 ± 0.61 | –0.05 ± 0.54 | .98 | ANOVA | |||
| No. of eyesa | 13 | % | 25 | % | 6 | % | – | – |
| Presenting VA 20/100 or better | 6 | 46 | 0 | 0 | 1 | 17 | .001 | Fisher |
| Presenting LP/NLP | 0 | 0 | 8 | 32 | 3 | 50 | .01 | Fisher |
| Final VA 20/100 or better | 7 | 54 | 1 | 4 | 1 | 17 | .001 | Fisher |
| Final LP/NLP | 1 | 8 | 8 | 32 | 4 | 67 | .02 | Fisher |
| Enucleated | 0 | 0 | 5 | 25 | 3 | 50 | – | Fisher |
Abbreviations: ANOVA, analysis of variance; LP, light perception; NLP, no light perception; VA, visual acuity.
a One eye with unknown etiology was not included in the data analysis.
Treatment and Outcomes
All patients were treated with IV and topical antibiotics. Initially, 23 (51%) eyes were treated with bedside intravitreal antibiotics; 13 (29%) eyes were treated with PPV and intravitreal antibiotics; 6 (13%) eyes were treated with IV and topical antibiotics only; and 3 (7%) eyes underwent primary enucleation. In total, 8 (18%) eyes ultimately were enucleated. Figure 1 depicts the course of treatment for all patients. One patient with bilateral severe Candida endophthalmitis had been treated elsewhere with systemic steroids for presumed noninfectious panuveitis for 3 months prior to presentation to us for progressive visual loss to HM vision. The steroids were discontinued, and she underwent bilateral PPV with intravitreal and systemic antifungal agents 2 times in 1 eye and 4 times in the other eye for progressive worsening infection and bilateral tractional retinal detachments. Another patient with methicillin-resistant Staphylococcus aureus EE secondary to endocarditis who was initially treated with PPV had 1 recurrence after finishing a protracted course of IV antibiotics. No systemic infection was found at the time of recurrence.
Figure 1.
Treatment course. IV indicates intravenous; PPV, pars plana vitrectomy.
*Refers to days following presentation.
At presentation, a view of the fundus was possible in 4 (66%) eyes, 7 (53%) eyes, and 9 (39%) eyes initially treated with IV antibiotics only, PPV with intravitreal antibiotics, and bedside intravitreal antibiotics, respectively. There was no significant difference (P = .44, Fisher exact test) in the treatment method selected when stratified by ability to view the fundus. When comparing all 3 initial treatment groups to each other, there was no significant difference in mean presenting VA (P = .07) (Table 5). Similarly, there was no significant difference in the proportion of eyes presenting with 20/100 or better VA, or those presenting with LP or worse, among the different treatment groups.
Table 5.
Visual Acuity by Initial Treatment Modality.
| Mean VA in logMAR | IV antibiotic only | Intravitreal + IV antibiotics | PPV with intravitreal + IV antibiotics | P | Test | |||
|---|---|---|---|---|---|---|---|---|
| Initial | 1.15 ± 0.81 | 1.80 ± 0.85 | 2.06 ± 0.57 | .07 | ANOVA | |||
| Final | 1.30 ± 0.99 | 1.83 ± 0.97 | 1.77 ± 0.93 | .55 | ANOVA | |||
| P, t test | .65 | .87 | .27 | – | – | |||
| Mean change | 0.15 ± 0.76 | –0.08 ± 0.67 | –0.29 ± 0.91 | .49 | ANOVA | |||
| No. of eyesa | 6 | % | 23 | % | 13 | % | – | – |
| Initial VA 20/100 or better | 3 | 50 | 3 | 13 | 1 | 8 | .06 | Fisher |
| Initial LP/NLP | 0 | 0 | 6 | 26 | 2 | 15 | .57 | Fisher |
| Final VA 20/100 or better | 2 | 33 | 5 | 38 | 2 | 15 | .67 | Fisher |
| Final LP/NLP | 0 | 0 | 8 | 35 | 2 | 15 | .17 | Fisher |
| Enucleation | 0 | 0 | 3 | 13 | 2 | 15 | – | |
Abbreviations: ANOVA, analysis of variance; IV, intravenous; LP, light perception; NLP, no light perception; PPV, pars plana vitrectomy; VA, visual acuity.
a Three eyes underwent primary enucleation.
At the last follow-up, the mean final VA across all eyes was 1.76 ± 0.98 logMAR or approximately 20/1100 at a mean follow-up time of 13 ± 24 months from presentation (median, 3 months). There was no significant difference in the mean final VA of eyes across the initial treatment groups. Table 5 describes visual outcomes as a function of initial treatment strategy.
On clearance of the infection, more eyes initially treated with PPV later required a second PPV (46%) as compared with those initially treated with IV antibiotics alone (17%) and those initially treated with bedside intravitreal antibiotics (13%), but this difference was not statistically significant (P = .08, Fisher exact test). Reasons for these second PPV included media opacities in 5 eyes (42%), tractional retinal detachment in 2 eyes (17%), rhegmatogenous retinal detachment in 1 eye (8%), combined rhegmatogenous/tractional retinal detachment in 3 eyes (25%), and vitreous hemorrhage in 1 eye (8%) (see Figure 1).
Four patients (11%) died during the inpatient admission from sepsis-related medical complications unrelated to the ocular infection. The VA and causative organisms at time of death of these 4 patients was 20/20 (Candida albicans), 20/40 (methicillin-sensitive S aureus), CF (Streptococcus agalactiae), and NLP (methicillin-resistant S aureus).
No eye that initially presented with VA of LP or NLP had any improvement in vision. Two eyes presenting with HM or better vision ultimately became NLP: 1 because of infection due to Aspergillus in a patient in the intensive care unit who was medically unstable to have surgery (20/30 at presentation) and 1 due to Serratia marcescens (HM at presentation).
Conclusions
EE is an uncommon complication of septicemia. The goal of this study was to report the treatment and outcomes of all patients with EE at our institution over a 13-year timeframe.
In North America and Europe, EE is caused primarily by gram-positive and fungal species, 1,6,7,10 -12 and the present study revealed similar findings; 56% and 29% of eyes with EE seen at our institution had gram-positive and fungal etiologies, respectively. On the other hand, EE in Asian countries is predominantly due to gram-negative species, 2,15 possibly because of the higher incidence of hepatobiliary pathology, which tends to be gram-negative in nature. 9 Correspondingly, 5 (82%) of the 6 eyes with gram-negative infection in our series had a hepatobiliary or genitourinary source of infection.
Nearly three-quarters of our patients had an identifiable systemic source of infection, with the most common sources being endocarditis (18%) and subcutaneous abscess (18%). This highlights the importance of complete physical examination and systemic workup in patients with EE, including echocardiogram and a thorough examination for skin ulcers or abscesses. Still, in a quarter of patients the complete systemic workup was negative for a source of infection; however, 90% of these patients had a history of IVDU.
Our study found that more than one-half of all patients with EE had a history of IVDU, which is consistent with recent reports, 3,6,11 but much greater than the 10% to 15% range reported 20 to 30 years ago. 8,10 IVDU often is thought of as a risk factor for fungal EE, 4,10 -12 perhaps because lemon juice used as a solvent to dissolve heroine in many cases promotes contamination with Candida, which grows very well in an acidic medium. 18,19 However, a recent series of patients with IVDU-associated EE showed that 16% of cases were, in fact, attributable to gram-positive organisms. 7 Of the patients in our series with IVDU-associated EE, 60% were bacterial whereas 40% were fungal in etiology. Thus, our results indicate that physicians should strongly consider fungal as well as bacterial etiologies when treating a patient with presumed EE and a history of IVDU.
A hypopyon was observed at presentation in 50% of our patients, which is higher than the 24% to 31% described in similar reports. 6,10 In our series, eyes with gram-positive infections were significantly more likely to present with hypopyon than eyes with fungal or gram-negative infections.
In this series, patients with fungal EE had positive systemic culture findings in only 36% of cases and were significantly less likely to have positive systemic culture results than those with gram-positive bacterial infections. The literature confirms similarly low rates of systemic culture positivity in eyes with fungal EE. 5,11,12
We found that eyes with fungal infection presented with better mean initial VA than eyes with bacterial infections. Although the mean final VA of eyes with fungal EE was not significantly different from those with bacterial infection, more fungal eyes had 20/100 or better final VA and fewer ultimately were LP or worse when compared with eyes with bacterial infections.
An important factor in determining visual prognosis in patients with EE may be the presenting VA. There was no significant difference in the mean change in VA, from baseline to last follow-up, when stratified by initial treatment method or by etiology of infection. Regarding visual prognosis, the literature shows mixed results. Modjtahedi et al, 6 in their series of patients with EE, demonstrated no difference between mean presenting VA and mean VA at week 2 and month 2 but did find a significant difference between mean presenting VA and mean VA at week 1 and month 6. They noted a 2-line or greater improvement in VA in 65% of their patients. Furthermore, in our series, no eye that presented with LP or worse recovered any vision with treatment. Similarly, a series by Okada and colleagues 8 described 13 eyes that presented with LP or worse vision and, at last follow-up, 12 of those eyes maintained their poor vision whereas only 1 improved to CF. Conversely, only 2 eyes that presented with HM or better vision progressed to LP/NLP vision.
Our results showed no significant difference in final visual outcome as a result of the initial treatment method. Similarly, there was no significant difference between the 3 initial treatment groups regarding the development of complications requiring a second PPV (ie, vitreous debris, retinal detachment, vitreous hemorrhage). This is in contrast to the series of Modjtahedi et al in which no eyes that underwent initial PPV required a second PPV as opposed to 55% of eyes that initially underwent bedside intravitreal antibiotics. 6 There were baseline differences in the populations between these 2 series that may underlie these different results, including a smaller number of eyes in our series. Different distributions of microbial etiologies, for example, might underlie the different outcomes. However, in our study, there was no difference in presenting VA or clarity of the fundus examination (indirectly assessing the severity of vitritis) among eyes initially treated differently, suggesting the different treatment groups were relatively well balanced in regard to the severity of infection.
Although we found no difference in visual outcome between initial treatment groups in our population, it is important to mention the retrospective, nonrandomized nature of this study. Treatment modalities were devised by the attending vitreoretina specialist using his or her clinical judgment to determine what was best for the patient, and thus, the lack of a difference in outcomes between initial treatment groups needs to be taken in appropriate context. Generally, if vitreous inflammation precludes a view of the optic nerve and macula, we usually intervene with PPV initially unless the patient’s unstable medical condition precludes surgery in the operating room.
This study was limited by its retrospective nature, its dependency on the accuracy of the medical record, and the relatively small number of patients evaluated. This series, albeit small, describes the geographic differences of EE in terms of the etiology and treatment modalities used in an inner-city tertiary care facility. A well-designed multicenter, prospective, randomized study or a meta-analysis of retrospective data may be a better way to compare the efficacies of various treatment regimens in eyes with EE.
Footnotes
Authors’ Note: A poster of this manuscript was presented at the 51st Retina Society Annual Meeting, in San Francisco, California, September 14, 2018.
Ethical Approval: Approval for this study was obtained from the Rutgers University Health Sciences Institutional Review Board and was conducted according to the Declaration of Helsinki.
Statement of Informed Consent: Informed consent was not required because the study was retrospective and presented data were not identifiable.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Greg Budoff, MD 
https://orcid.org/0000-0002-2502-5226
References
- 1. Solborg Bjerrum S, la Cour M. 59 eyes with endogenous endophthalmitis—causes, outcomes and mortality in a Danish population between 2000 and 2016. Graefes Arch Clin Exp Ophthalmol. 2017;255(10):2023–2027. doi:10.1007/s00417-017-3760-4 [DOI] [PubMed] [Google Scholar]
- 2. Chung KS, Kim YK, Song YG, et al. Clinical review of endogenous endophthalmitis in Korea: a 14-year review of culture positive cases of two large hospitals. Yonsei Med J. 2011;52(4):630–634. doi:10.3349/ymj.2011.52.4.630 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Connell PP, O’Neill EC, Fabinyi D, et al. Endogenous endophthalmitis: 10-year experience at a tertiary referral centre. Eye (Lond). 2011;25(1):66–72. doi:10.1038/eye.2010.145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Connell PP, O'Neill EC, Amirul Islam FM, et al. Endogenous endophthalmitis associated with intravenous drug abuse: seven-year experience at a tertiary referral center. Retina. 2010;30(10):1721–1725. doi:10.1097/IAE.0b013e3181dd6db6 [DOI] [PubMed] [Google Scholar]
- 5. Lingappan A, Wykoff CC, Albini TA, et al. Endogenous fungal endophthalmitis: causative organisms, management strategies, and visual acuity outcomes. Am J Ophthalmol. 2012;153(1):162–166.e1. doi:10.1016/j.ajo.2011.06.020 [DOI] [PubMed] [Google Scholar]
- 6. Modjtahedi BS, Finn AP, Barb SM, et al. Characteristics and outcomes of endogenous endophthalmitis. Ophthalmol Retina. 2019;3(1):61–72. doi:10.1016/j.oret.2018.08.009 [DOI] [PubMed] [Google Scholar]
- 7. Modjtahedi BS, Finn AV, Papakostas TD, Durand M, Husain D, Eliott D. Intravenous drug use–associated endophthalmitis. Ophthalmol Retina. 2017;1(3):192–199. doi:10.1016/j.oret.2016.10.013 [DOI] [PubMed] [Google Scholar]
- 8. Okada AA, Johnson RP, Liles WC, D’Amico DJ, Baker AS. Endogenous bacterial endophthalmitis: report of a ten-year retrospective study. Ophthalmology. 1994;101(5):832–838. doi:10.1016/S0161-6420(13)31255-X [PubMed] [Google Scholar]
- 9. Park IH, Jun CH, Wi JW, et al. Prevalence of and risk factors for endogenous endophthalmitis in patients with pyogenic liver abscesses. Korean J Intern Med. 2015;30(4):453–459. doi:10.3904/kjim.2015.30.4.453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Schiedler V, Scott IU, Flynn HW, Jr, Davis JL, Benz MS, Miller D. Culture-proven endogenous endophthalmitis: clinical features and visual acuity outcomes. Am J Ophthalmol. 2004;137(4):725–731. doi:10.1016/j.ajo.2003.11.013 [DOI] [PubMed] [Google Scholar]
- 11. Sridhar J, Flynn HW, Jr, Kuriyan AE, Miller D, Albini T. Endogenous fungal endophthalmitis: risk factors, clinical features, and treatment outcomes in mold and yeast infections. J Ophthalmic Inflamm Infect. 2013;3(1):60. doi:10.1186/1869-5760-3-60 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Tirpack AR, Duker JS, Baumal CR. An outbreak of endogenous fungal endophthalmitis among intravenous drug abusers in New England. JAMA Ophthalmol. 2017;135(6):534–540. doi:10.1001/jamaophthalmol.2017.0650 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Vaziri K, Pershing S, Albini TA, Moshfeghi DM, Moshfeghi AA. Risk factors predictive of endogenous endophthalmitis among hospitalized patients with hematogenous infections in the United States. Am J Ophthalmol. 2015;159(3):498–504. doi:10.1016/j.ajo.2014.11.032 [DOI] [PubMed] [Google Scholar]
- 14. Budoff G, Thangamathesvaran L, Zarbin MA, Bhagat N. Bacterial endogenous endophthalmitis in bacteremic inpatients. Ophthalmol Retina. 2019;3(11):971–978. doi:10.1016/j.oret.2019.06.004 [DOI] [PubMed] [Google Scholar]
- 15. Wong JS, Chan TK, Lee HM, Chee SP. Endogenous bacterial endophthalmitis: an East Asian experience and a reappraisal of a severe ocular affliction. Ophthalmology. 2000;107(8):1483–1491. doi:10.1016/s0161-6420(00)00216-5 [DOI] [PubMed] [Google Scholar]
- 16. Holladay JT. Proper method for calculating average visual acuity. J Refract Surg. 1997;13(4):388–391. doi:10.3928/1081-597X-19970701-16 [DOI] [PubMed] [Google Scholar]
- 17. The IONDT Research Group. The Ischemic Optic Neuropathy Decompression Trial (IONDT): design and methods. Control Clin Trials. 1998;19(3):276–296. doi:10.1016/s0197-2456(98)00003-8 [DOI] [PubMed] [Google Scholar]
- 18. Shankland G, Richardson M. Possible role of preserved lemon juice in the epidemiology of Candida endophthalmitis in heroin addicts. Eur J Clin Microbiol Infect Dis. 1989;8(1):87–89. doi:10.1007/BF01964128 [DOI] [PubMed] [Google Scholar]
- 19. Albini TA, Sun RL, Holz ER, Khurana RN, Rao NA. Lemon juice and Candida endophthalmitis in crack-cocaine misuse. Br J Ophthalmol. 2007;91(5):702–703. doi:10.1136/bjo.2006.0108365 [DOI] [PMC free article] [PubMed] [Google Scholar]