Chapter 169 – Endophthalmitis
RUSSELL W. READ
• Inflammation within the anterior or posterior segment, or both, traditionally occurs with concurrent partial-thickness involvement of an adjacent ocular wall. Endophthalmitis may be either infectious or noninfectious.
• Decreased vision.
• Anterior chamber or vitreous cellular reaction.
• Conjunctival hyperemia.
• Lid edema.
Endophthalmitis is a potentially severe intraocular inflammation which may occur as a complication of intraocular surgery or as a result of nonsurgical trauma or systemic infection. Endophthalmitis is not necessarily an infectious process, therefore lens-induced inflammation and severe noninfectious postoperative inflammations may be termed sterile endophthalmitis. In common usage, however, the term endophthalmitis indicates an infectious cause.
EPIDEMIOLOGY AND PATHOGENESIS
Endophthalmitis can be classified as in Box 169-1 . Organisms causing infectious endophthalmitis usually originate from an exogenous source, entering the eye following intraocular surgery including cataract extraction, secondary lens implantation, pars plana vitrectomy, glaucoma filter, penetrating keratoplasty, and others. Infection may also follow a penetrating ocular injury or occur as a result of hematogenous spread from a systemic infection.
Acute Infectious Postoperative Endophthalmitis
Acute endophthalmitis typically is defined as occurring within 6 weeks of surgery. The incidence of endophthalmitis following intraocular surgery has decreased steadily during the twentieth century. Studies published within the last 10 years have reported an incidence of acute endophthalmitis following cataract extraction of between 0.072% and 0.13%.    
The most common infecting organisms following cataract extraction are the coagulase-negative Staphylococcus spp., especially S. epidermidis.  Gram-negative bacteria and anaerobes are much less frequent causative agents. The most likely source of infectious agents is the patient’s own lid and conjunctival flora, with entry at the time of surgery. One study showed a 29% incidence of positive aqueous culture results taken at the time of cataract extraction. Because the incidence of endophthalmitis is much lower than 29%, factors other than the presence of bacteria in the eye must be at play. Infection may occur due to an overwhelming of intrinsic ocular defense mechanisms such as clearance of organisms by aqueous outflow, complement activation, and phagocytosis. Preoperative risk factors include active ocular surface infections or colonization, such as blepharitis, conjunctivitis, lacrimal drainage system infection or obstruction, and contaminated eyedrops. Operative risks include wound abnormalities, vitreous loss, prolonged surgery, and contaminated irrigation solutions. Thus careful preoperative evaluation and correction of preexisting risk factors, application of topical povidone-iodine (5%) solution prior to surgery, careful draping to isolate the lid margin and lashes from the surgical wound, and careful surgical technique should reduce the risk of endophthalmitis. The issue of perioperative antibiotic use is one of great controversy and is yet to be resolved.
Delayed-Onset Infectious Endophthalmitis
Delayed-onset endophthalmitis is defined as that occurring more than 6 weeks following surgery. Persistent or recurrent uveitis following cataract extraction is difficult to differentiate from a smoldering or delayed-onset postoperative infection. Thus the clinician must have a high index of suspicion. Typically organisms of lower virulence, such as Propionibacterium acnes, S. epidermidis, and fungi are involved.
Bleb-Associated Infectious Endophthalmitis
Bleb-associated endophthalmitis following a glaucoma filtering procedure may range from blebitis to frank purulent endophthalmitis and may occur during the early or late postoperative periods. The incidence of acute infectious endophthalmitis following glaucoma filtering surgery has been reported to be 0.061–0.3%.  The most common causative organisms are Streptococcus spp. The reported incidence of delayed bleb-related endophthalmitis is in the range of 0.2–18%.  The most common causative organisms may differ from those found in acute-onset disease and include Streptococcus spp. and Haemophilus influenzae.  Local antimetabolite adjunctive therapy further increases the risk of bleb-related endophthalmitis.
Other Intraocular Surgery
Endophthalmitis also can occur, but rarely, following pars plana vitrectomy, penetrating keratoplasty, and pneumatic retinopexy. Corneal transplantation presents a unique opportunity for inoculation of microorganisms via the donor graft tissue. Thus most surgeons submit the donor rim for culture at the time of surgery.
Posttraumatic Infectious Endophthalmitis
The incidence of endophthalmitis following penetrating ocular trauma is approximately 7%, but it may be as high as 30% in
Classification of Endophthalmitis
rural locales. Presence of an intraocular foreign body increases the risk. Although gram-positive cocci are still the most common isolates,  Bacillus spp. and other virulent organisms, sometimes polymicrobial, are potential causes.
Endogenous endophthalmitis occurs as a result of the hematogenous or direct spread of microorganisms from a site external to the eye and mandates a thorough systemic search for the site of origin. One series reported that organisms causing endocarditis and those in the gastrointestinal tract were the most common primary sources. Organisms may be those common to other types of endophthalmitis, such as streptococcal and staphylococcal spp., or unusual organisms, including fungi.
Acute postoperative endophthalmitis by definition occurs within 6 weeks after surgery. Severity may range from mild to severe. Milder forms—usually due to less virulent organisms—tend to become apparent later than the more severe form, which usually occurs within the first 6 postoperative days. The most common symptoms reported in the Endophthalmitis Vitrectomy Study (EVS) were blurred vision, occurring in 94% of patients; a red eye, occurring in 82% of patients; and pain, reported by 74% of patients.
Critical findings on examination include decreased visual acuity, eyelid edema and erythema, conjunctival hyperemia and chemosis, corneal edema and opacification, anterior chamber cell and flare (more than expected for a typical postoperative course) frequently with a hypopyon (found in 86% of patients in the EVS ) ( Fig. 169-1 ), vitritis, and scattered retinal hemorrhages and periphlebitis if the retina is visible. In 79% of EVS patients, no view of the retinal vessels was possible with indirect ophthalmoscopy, and a red reflex was present in only 32% of patients.
In delayed-onset endophthalmitis, the clinical picture is frequently indistinguishable from that of anterior uveitis. Patients may complain of photophobia, blurred vision, and mild pain. Keratic precipitates with anterior chamber and vitreous cells and flare can be seen. A capsular plaque is very typical in cases of P. acnes endophthalmitis ( Fig. 169-2 ). Endophthalmitis may develop following Nd:YAG laser capsulotomy, presumably due to release of previously sequestered low-virulence organisms into the vitreous.
Endophthalmitis associated with filtering blebs can occur anytime after surgery. Symptoms and signs are similar to those seen with post-cataract extraction endophthalmitis, with the addition of a hypopyon or cellular debris within the bleb. Patients may exhibit the latter signs without vitreous involvement, termed blebitis.
Figure 169-1 Hypopyon following cataract extraction. Note the presence of neovascularization of the iris.
Figure 169-2 Propionibacterium acnes plaque on the posterior capsule following cataract extraction and posterior chamber lens insertion. (Courtesy of Howard H. Tessler.)
In posttraumatic endophthalmitis, the onset may occur anytime from days to weeks following injury. Diagnosis is not uncommonly delayed due to the difficulty inherent in differentiating expected posttraumatic inflammation from infection. As mentioned previously, an intraocular foreign body increases the risk of infection.
Patients who have endogenous endophthalmitis most frequently experience decreased vision and floaters in one or both eyes. They typically experience less inflammation and pain than those with other forms of endophthalmitis. Retinal, subretinal, and choroidal infiltrates are seen ( Fig. 169-3 ). Depending on the timing of diagnosis after onset and degree of immunosuppression present, the vitreous may contain cells and debris overlying these infiltrates. Involvement of both eyes occurs in one fourth of cases, but involvement of the fellow eye may be delayed several days to weeks.
DIAGNOSIS AND ANCILLARY TESTING
Early recognition is critical; therefore a high index of suspicion must be maintained. A complete ocular and medical history is taken and a thorough ophthalmic examination performed. Ultrasonography is helpful, especially in cases with significant anterior-segment media opacity, to confirm the presence of vitreous cells, detect a retinal or choroidal detachment or both, and to search for retained lens remnants in the posterior segment. An
Figure 169-3 Candidal retinal infiltrate in an intravenous drug abuser. Note the small exudative detachment adjacent to the infiltrate.
“A” scan is helpful to distinguish between vitreal membranes and retinal detachment. The combination of hypopyon and echographically clear vitreous may be suggestive of an early coagulase-negative staphylococcal or culture-negative endophthalmitis. An ultrasonographic examination also is important as a baseline against which the success of treatment can be measured.
Although therapy should not be delayed awaiting culture results, diagnostic procedures are necessary to obtain intraocular fluid to guide subsequent therapy and to confirm that adequate coverage was achieved with empirical therapy. Use of systemic or topical antibiotics prior to diagnostic procedures may decrease the yield of culturable organisms from intraocular fluid sample collections. An anterior chamber paracentesis is performed using either a 25- or 27-gauge needle and approximately 0.1?ml of aqueous material aspirated. Vitreous material can be obtained via several methods. A trans–pars plana aspiration may be performed, using a 23-gauge needle inserted into the anterior vitreous cavity, entering 3?mm posterior to the limbus in pseudophakic or aphakic eyes and 4?mm posterior to the limbus in phakic eyes. If possible, the needle should be visualized in the vitreous cavity. Approximately 0.2?ml of liquid vitreous is aspirated. Small and careful movements of the needle in the vitreous cavity may be necessary to obtain an adequate sample. Small-gauge battery-powered vitreous cutting instruments are available for office use and may decrease the risk of an iatrogenic retinal tear by cutting, rather than pulling, on the vitreous body. Finally, a traditional three-port vitrectomy can be performed. The latter has the advantages of production of a larger sample volume, debulking the vitreous of toxic products, decreasing the infecting agent load, and releasing possible sources of vitreous traction. Comparison between immediate vitrectomy and vitreous sampling, as regards visual acuity and media clarity, was a major goal of the Endophthalmitis Vitreous Study (EVS). This study found that immediate vitrectomy produced a better outcome only in eyes with light perception vision at the time of evaluation.
Aqueous and vitreous samples should be plated on blood agar, chocolate agar, Sabouraud dextrose agar, thioglycollate broth, and anaerobic medium. Gram and Giemsa stains should be performed also. The remainder of the specimen should be mixed with an equal volume of 95% alcohol and submitted for pathological evaluation. The polymerase chain reaction may provide a more rapid method of specific diagnosis by allowing detection of DNA from infecting organisms. Issues with contamination and ability to provide rapid speciation of the amplified DNA products need to be resolved before this methodology can totally supplant culture. Traditional culture results are generally positive within 48 hours, with sensitivities usually available 24 hours later. The EVS yielded a 68.2% incidence of confirmed
Differental Diagnosis of Endophthalmitis
History of intraocular surgery or trauma
Severe sterile postoperative inflammation
Retained lens nuclear material
History of or predisposition to uveitis
Inadvertent toxic substance introduced into eye
Metipranolol (usually mild)
Latanoprost (usually mild)
Rifabutin (frequently with hypopyon)
Rebound inflammation after rapid antiinflammatory taper
No history of intraocular surgery or trauma
positive culture results, similar to those found in other series. Reasons for negative culture results include fastidious organisms, insufficient sampling, and “sterile” endophthalmitis.
The differential diagnosis of endophthalmitis is detailed in Box 169-2 .
Endogenous endophthalmitis typically occurs with sepsis, an immunocompromised state (such as by the acquired immunodeficiency syndrome; immunosuppressive therapy, including corticosteroid use; long-term antibiotic use; or disseminated cancer), presence of an indwelling catheter (urethral or intravenous), or intravenous drug abuse. Patients with these conditions usually do not have a history of recent intraocular surgery.
Acute endophthalmitis is characterized by the influx of neutrophils. The inflammatory process may be suppurative, as determined by the production of pus, or nonsuppurative. Certain bacteria, such as the Staphylococci, have a propensity to produce purulent inflammation and are termed pyogenic. Necrosis of the involved ocular structures is common. Neutrophils actively produce reactive oxygen species as one mechanism of host defense, and these toxic substances cause much of the damage and subsequent visual loss that occurs in infectious endophthalmitis. Repair of damaged tissues results in scar formation via the mechanisms of fibrosis and gliosis.
Chronic inflammation is characterized by the presence of lymphocytes and is divided further into granulomatous and nongranulomatous forms. Granulomatous inflammation is characterized by the presence of epithelioid cells and giant cells and may occur with fungal infection.
Acute infectious endophthalmitis is a true ophthalmic emergency and mandates prompt therapy if visual acuity is to be preserved.
Because rapid administration of antibiotics is necessary, waiting for culture results or even a Gram stain is not feasible. Although most studies, including the EVS, have shown a majority of isolates
Empirical Medical Therapy of Endophthalmitis
ACUTE ONSET POST–CATARACT EXTRACTION 
Vancomycin hydrochloride 1.0?mg in 0.1?ml (normal saline) and
Ceftazidime 2.25?mg in 0.1?ml (normal saline) or Amikacin 200–400?µg in 0.1?ml (normal saline)
Dexamethasone 400?µg in 0.1?ml (optional)
Vancomycin hydrochloride 25?mg in 0.5?ml (normal saline) and
Ceftazidime 100?mg in 0.5?ml (normal saline) or Amikacin 25?mg in 0.5?ml (normal saline) if ß-lactam allergy exists and
Dexamethasone 6?mg in 0.25?ml (normal saline)
Vancomycin hydrochloride 50?mg/ml and
Amikacin 20?mg/ml and
Atropine sulfate 1% or scopolamine hydrobromide 0.25% and
Prednisolone acetate 1%
Prednisone 30?mg twice daily for 5–10 days (optional)
Parallel to those listed for post–cataract extraction, and in addition:
May also consider use of intravitreal clindamycin phosphate (450?µg)
Systemic antibiotics still considered standard of care, options include selections from the following:
Clindamycin 600–900?mg intravenously every 8 hours
Ceftazidime 2?g intravenously every 8 hours
Amikacin 7.5?mg/kg intravenously once, then 6?mg/kg every 12 hours
Ciprofloxacin 750?mg po twice daily
May respond to above topical regimen alone
Parallel to post–cataract extraction, but consider addition of systemic antibiotics as well
from postsurgical and posttraumatic infections are gram positive, empirical broad spectrum coverage for both gram-positive and gram-negative organisms is necessary. Box 169-3 lists antibiotics and dosages. Vancomycin is widely accepted as the intravitreous agent of choice for gram-positive coverage. Studies have shown that a dose of 1.0?mg is well tolerated and nontoxic in a rabbit model. In several series of endophthalmitis, 100% of gram-positive organisms cultured were found to be sensitive to vancomycin. Concerns over resistance developing are probably not realistic when used intravitreally, in sharp contrast to the legitimate concern when a valuable agent such as vancomycin is used indiscriminately as a prophylactic agent.
Gram-negative intravitreous coverage may be provided with a variety of agents, including aminoglycosides and ß-lactams. Each has advantages and disadvantages. Aminoglycosides have a synergistic effect with vancomycin for the treatment of enterococci. Aminoglycosides, however, may produce macular infarction. Amikacin appears to present less risk than other aminoglycosides, especially gentamicin. Third-generation cephalosporins, such as ceftazidime 2.25?mg, may be used instead, and they provide good coverage for gram-negative organisms without the risk of retinal damage. If aminoglycosides are used, their repeated intravitreal injection over a several-day period should be avoided.
Although gram-positive cocci are the most common cause of posttraumatic endophthalmitis, the risk of infection with organisms such as Bacillus must be considered. One study from India showed an 84% sensitivity rate of Bacillus to vancomycin, 43% sensitivity rate to ceftazidime, and a 100% sensitivity rate to amikacin, gentamicin, and ciprofloxacin.  Another study, also from India, showed an 87% sensitivity to amikacin, 94% sensitivity to gentamicin, and 68% sensitivity to vancomycin. Ciprofloxacin appears to have reasonably good activity against Bacillus, and studies have shown adequate penetration into the vitreous after a single 750?mg oral dose. Other options include clindamycin (see Box 169-3 ).
Fungal endophthalmitis is treated with intravitreal amphotericin B (5?mg in 0.1?ml) after a positive culture result is obtained or if there is reason for strong suspicion of fungal infection. Systemic therapy also should be given, but renal toxicity must be monitored closely. Adjunctive corticosteroid therapy should not be given for eyes in which fungal endophthalmitis is suspected.
Clinical studies in bacterial endophthalmitis have reported varied results as regards the use of concurrent intravitreal corticosteroids. In a retrospective review, visual acuity and inflammation were improved in eyes injected with intravitreous dexamethasone. However, in a prospective, randomized trial evaluating vitrectomy with coadministration of intravitreal dexamethasone and antibiotics in exogenous bacterial endophthalmitis (including both postoperative and posttraumatic causes), Das et al. found that intravitreal dexamethasone reduced early inflammatory scores but had no influence on the visual outcome as measured at 12 weeks. Because the EVS did not randomize patients to receive glucocorticoids or not, it does not address this issue.
Subconjunctival injection of dexamethasone (12?mg) is used commonly, as are topical glucocorticoids. Systemic glucocorticoids can be administered orally (30?mg twice a day for 5–10 days ) if there are no contraindications.
The role of systemic antibiotic therapy had been controversial in postoperative endophthalmitis, but the EVS reported no difference in final visual acuity and media clarity with or without the use of systemic antibiotics for acute post–cataract extraction endophthalmitis. The role of systemic antibiotics in other forms of endophthalmitis has not been addressed, and many clinicians still consider their use the standard of care in these situations.  Systemic antimicrobial usage is mandatory in endogenous endophthalmitis and is tailored to the offending organism. Infectious disease consultation is recommended both for the systemic work-up and systemic treatment of endogenous endophthalmitis. Intravitreal antibiotics are required if the infecting agent has gained access to the vitreous cavity or empirically if any question exists.
In one sense, surgical management of endophthalmitis begins before the infection occurs. Careful operative technique to minimize wound abnormalities, avoidance of vitreous loss during cataract surgery, and careful microsurgical wound management and closure in open globe injuries decrease the risk. Once endophthalmitis has begun, early options primarily center around the method of intraocular fluid sampling and response to therapy ( Fig. 169-4 ).
Later surgery to deal with complications may also be required. The advantages of early therapeutic vitrectomy are clearing of the ocular media, removal of potentially harmful bacterial products, reduction of bacterial load, and removal of the vitreous scaffolding by which traction retinal detachments may occur. Disadvantages include delay in treatment until operating room time is available, iatrogenic retinal holes or detachments, choroidal hemorrhage, and the problem of visualizing the posterior segment in an eye that has had recent surgery. Retinal detachment is difficult to treat in eyes that undergo vitrectomy for endophthalmitis due to the need for air-fluid exchange and injection of aqueous antibiotic. Concentration of antibiotic in the aqueous layer may lead to an increased risk of toxicity. In acute-onset post–cataract extraction endophthalmitis, the EVS showed that no difference exists in visual outcome for eyes with visual acuity of hand motions or better with or without vitrectomy. In the subgroup of patients who have initial light perception, vitrectomy produced a threefold increase in the frequency
Figure 169-4 Algorithm for management of acute endophthalmitis after cataract extraction.
of achieving 20/40 (6/12) vision or better, a twofold increased chance of achieving 20/100 (6/30) vision or better, and a 50% decrease in the frequency of severe visual loss. In addition, no difference was found in the confirmed culture results between the group of patients who had undergone tap versus those with a therapeutic vitrectomy.
Chronic endophthalmitis caused by P. acnes frequently requires surgery, not only to confirm the diagnosis but also to remove any sequestered infectious material from the posterior capsule, along with injection of intravitreous antibiotics, usually vancomycin. If this is not successful, removal of the intraocular lens with complete en bloc removal of the anterior and posterior capsules may be necessary.
Endogenous endophthalmitis may require surgical intervention in the form of vitreal, retinal, and choroidal biopsies, and culture if there is no obvious primary source and blood cultures and other studies are negative.
COURSE AND OUTCOME
The findings from the EVS that patients with post–cataract surgery endophthalmitis with hand-motion or better acuity had no benefit from either systemic antibiotic therapy or immediate vitrectomy have moved management of this specific category of patients primarily to the outpatient office setting. Regardless, the patient should be monitored on a daily basis. Admission should be considered if a patient cannot return on a frequent basis, has a history of trauma, or is believed to have endogenous endophthalmitis. Cultures usually require 24–48 hours for initial results and need to be checked daily.
The axiom “if it isn’t worse, it’s better,” may apply, because media clarity and visual acuity may not improve initially. Level of pain and lid injection may be helpful in determining an early response. Repeat intravitreous injection of antibiotics may be required if the condition worsens and infection persists as confirmed by repeat culture. Serial ultrasonography may be used to monitor clinical response and detect retinal detachment.
Baseline Risk Factors for Decreased Visual Acuity Outcome
Corneal infiltrate or ring ulcer
Posterior capsule not intact
Intraocular pressure less than 5?mm Hg or greater than 25?mm Hg
Afferent pupillary defect
Absent red reflex
Visual acuity of light perception, the most important risk factor, with a twofold greater risk of poor visual outcome compared with those with hand-motion or better acuity during initial evaluation
Risk factors for decreased visual acuity outcome found at baseline examination, as determined by the EVS, are detailed in Box 169-4 . EVS data showed that 53.1% of patients had a final visual acuity of >20/40 (6/12), 74.4% of >20/100 (6/30), and 88.6% of >5/200 (6/240).
1. Aaberg TM Jr, Flynn HW Jr, Schiffman J, Newton J. Nosocomial acute-onset postoperative endophthalmitis survey. A 10-yr review of incidence and outcomes. Ophthalmology. 1998;105:1004–10.
2. Powe NR, Schein OD, Gieser SC, et al. Synthesis of the literature on visual acuity and complications after cataract extraction with intraocular lens implantation. Arch Ophthalmol. 1994;112:239–52.
3. Javitt JC, Vitale S, Canner JK, et al. National outcomes of cataract extraction: endophthalmitis after inpatient surgery. Arch Ophthalmol. 1991;109:1085–9.
4. Kattan HM, Flynn HW Jr, Pflugfelder SC, et al. Nosocomial endophthalmitis survey. Current incidence of infection after intraocular surgery. Ophthalmology. 1991;98:227–38.
5. Han DP, Wisniewski SR, Wilson LA, et al. Spectrum and susceptibilities of microbiologic isolates in the Endophthalmitis Vitrectomy Study. Am J Ophthalmol. 1996;122:1–17 [erratum in Am J Ophthalmol. 1996;122:920].
6. Kunimoto DY, Das T, Sharma S, et al. Microbiologic spectrum and susceptibility of isolates: part I. Postoperative endophthalmitis. Endophthalmitis Research Group. Am J Ophthalmol. 1999;128:240–2.
7. Beigi B, Westlake W, Mangelschots E, et al. Perioperative microbial contamination of anterior chamber aspirates during extracapsular cataract extraction and phacoemulsification. Br J Ophthalmol. 1997;81:953–5.
8. Maxwell DP, Brent BD, Orillac R, et al. A natural history study of experimental Staphylococcus epidermidis endophthalmitis. Curr Eye Res. 1993;12:907–12.
9. Driebe WT Jr, Mandelbaum S, Forster RK, et al. Pseudophakic endophthalmitis. Diagnosis and management. Ophthalmology. 1986;93:442–8.
10. Bohigian GM. A study of the incidence of culture-positive endophthalmitis after cataract surgery in an ambulatory care center. Ophthalmic Surg Lasers. 1999;30:295–8.
11. Ciulla TA, Beck AD, Topping TM, Baker AS. Blebitis, early endophthalmitis, and late endophthalmitis after glaucoma-filtering surgery. Ophthalmology. 1997;104: 986–95.
12. Mandelbaum S, Forster RK, Gelender H, Culbertson W. Late onset endophthalmitis associated with filtering blebs. Ophthalmology. 1985;92:964–72.
13. Waheed S, Ritterband DC, Greenfield DS, et al. New patterns of infecting organisms in late bleb-related endophthalmitis: a ten year review. Eye. 1998;12:910–5.
14. Kangas TA, Greenfield DS, Flynn HW Jr, et al. Delayed-onset endophthalmitis associated with conjunctival filtering blebs. Ophthalmology. 1997;104:746–52.
15. Greenfield DS, Suner IJ, Miller MP et al. Endophthalmitis after filtering surgery with mitomycin. Arch Ophthalmol. 1996;114:943–9.
16. Hilton GF, Tornambe PE. Pneumatic retinopexy. An analysis of intraoperative and postoperative complications. The Retinal Detachment Study Group. Retina. 1991;11:285–94.
17. Merchant A, Zacks CM, Wilhelmus K, et al. Candidal endophthalmitis after keratoplasty. Cornea. 2001;20:226–9.
18. Kresloff MS, Castellarin AA, Zarbin MA. Endophthalmitis. Surv Ophthalmol. 1998;43:193–224.
19. Boldt HC, Pulido JS, Blodi CF, et al. Rural endophthalmitis. Ophthalmology. 1989;96:1722–6.
20. Kunimoto DY, Das T, Sharma S, et al. Microbiologic spectrum and susceptibility of isolates: part II. Posttraumatic endophthalmitis. Endophthalmitis Research Group. Am J Ophthalmol. 1999;128:242–4.
21. Abu el-Asrar AM, al-Amro SA, al-Mosallam AA, al-Obeidan S. Post-traumatic endophthalmitis: causative organisms and visual outcome. Eur J Ophthalmol. 1999;9:21–31.
22. Duch-Samper AM, Chaques-Alepuz V, Menezo JL, Hurtado-Sarrio M. Endophthalmitis following open-globe injuries. Curr Opin Ophthalmol. 1998; 9:59–65.
23. Okada AA, Johnson RP, Liles WC, et al. Endogenous bacterial endophthalmitis. Report of a ten-year retrospective study. Ophthalmology. 1994;101:832–8.
24. Essman TF, Flynn HW Jr, Smiddy WE, et al. Treatment outcomes in a 10-year study of endogenous fungal endophthalmitis. Ophthalmic Surg Lasers. 1997;28: 185–94.
25. Endophthalmitis Vitrectomy Study Group. Results of the endophthalmitis vitrectomy study: a randomized trial of immediate vitrectomy and of intravenous antibiotics for the treatment of postoperative bacterial endophthalmitis. Arch Ophthalmol. 1995;113:1479–96.
26. Carlson AN, Koch DD. Endophthalmitis following Nd:YAG laser posterior capsulotomy. Ophthalmic Surg. 1988;19:168–70.
27. Dacey MP, Valencia M, Lee MB, et al. Echographic findings in infectious endophthalmitis. Arch Ophthalmol. 1994;112:1325–33.
28. Keren G, Alhalel A, Bartov E, et al. The intravitreal penetration of orally administered ciprofloxacin in humans. Invest Ophthalmol Vis Sci. 1991;32:2388–92.
29. Mao LK, Flynn HW Jr, Miller D, Pflugfelder SC. Endophthalmitis caused by Staphylococcus aureus. Am J Ophthalmol. 1993;116:584–9.
30. Das T, Jalali S, Gothwal VK, et al. Intravitreal dexamethasone in exogenous bacterial endophthalmitis: results of a prospective randomised study. Br J Ophthalmol. 1999;83:1050–5.
31. Sternberg P Jr, Martin DF. Management of endophthalmitis in the post-endophthalmitis vitrectomy study era. Arch Ophthalmol. 2001;119:754–5.
32. Bhagat N, Read RW, Rao NA, et al. Rifabutin-associated hypopyon uveitis in human immunodeficiency virus–negative immunocompetent individuals. Ophthalmology. 2001;108:750–2.
33. Aaberg TM Jr, Sternberg P. Trauma: principles and techniques of treatment. In: Ryan SJ, ed. Retina, vol 3, ed 3. St. Louis: Mosby; 2001:2400–26.
34. Doft BH. The endophthalmitis vitrectomy study. In: Kertes PJ, Conway MD, eds. Clinical trials in ophthalmology: a summary and practice guide. Philadelphia: Lippincott, Williams & Wilkins; 1998:97–111.