Richard K. Forster

Nosocomial Endophthalmitis Survey: Current Incidence of Infection after Intraocular Surgery

The authors reviewed the incidence of hospital-linked postoperative endophthalmitis at the Bascom Palmer Eye Institute between January 1, 1984 and June 30, 1989. After 30,002 intraocular surgical procedures, the following incidence of culture-proven endophthalmitis was observed: (1) extracapsular cataract extraction (ECCE) with or without intraocular lens (IOL) implantation--0.072% (17 of 23,625 cases); (2) pars plana vitrectomy--0.051% (1 of 1974 cases); (3) penetrating keratoplasty (PKP)--0.11% (2 of 1783 cases); (4) secondary IOL--0.30% (3 of 988 cases); and (5) glaucoma filtering surgery--0.061% (1 of 1632 cases). A statistically significant (P = 0.038, Fishers exact test, two-tailed) increased incidence of endophthalmitis occurred in diabetic (0.163%, 6 of 3686 cases) compared with nondiabetic (0.055%, 11 of 19,939 cases) patients undergoing ECCE with or without IOL implantation. The authors also reviewed the incidence of postoperative endophthalmitis after intracapsular cataract extraction (ICCE) with and without IOL and observed an incidence of 0.093% (7 of 7552) in cases operated on between September 1, 1976 and December 31, 1982.

Spectrum of Microbial Keratitis in South Florida

During a nine-year period ending in 1977, we scraped and cultured 663 corneal ulcers. Of these cases of keratitis, 238 were bacterial infections, 133 were fungal, and 292 were culture-negative. Pseudomonas was the predominant bacterial organism, and Fusarium was the most common fungus isolated. November was the peak month for both bacterial and fungal keratitis. Direct inoculation of multiple media, including Sabourauds agar, blood agar, chocolate agar, thioglycollate liquid, and brain-heart infusion liquid, enhanced the recovery rate; each medium provides special nutrients for different organisms. The Gram and Giemsa stains were satisfactory, but the newer Grocott methenamine silver stain for fungi and the limulus lysate test for gram-negative bacteria proved to be clinically useful. The clinical profile of patients with fungal keratitis differed from that of patients with Pseudomonas keratitis. Thirty-three percent of the patients with Pseudomonas keratitis were wearing contact lenses at the time of their infection, and the remainder had a high incidence of predisposing ocular conditions. Fungal keratitis tended to occur in healthy male patients who had been subjected to outdoor trauma.

Management of Infectious Endophthalmitis

Diagnostic anterior chamber and vitreous aspiration confirmed an infectious etiology in 78 of 140 eyes (56%) with suspected endophthalmitis. In 27 eyes the vitreous aspirate was positive, while the anterior chamber aspirate was negative. Intraocular antibiotics were used in 88 eyes including 50 which underwent therapeutic vitrectomy. Vision of 20/20 to 20/400 was achieved in 57% of recently operated, culture-positive eyes treated with intraocular antibiotics, and in 59% of those treated with combined vitrectomy and intraocular antibiotics.

Pseudophakic endophthalmitis: diagnosis and management

Eighty-three cases of endophthalmitis occurring in eyes with intraocular lenses are presented. A problem with the cataract section was felt to contribute to development of the endophthalmitis in 22% of patients. Seventy-five percent of eyes were culture positive with a microbiologic spectrum similar to other reported series without lens implants. Staphylococcus epidermidis was the most common organism accounting for 38% of the isolates. Therapeutic vitrectomy was performed in 46 cases (55%), generally in those eyes with the most severe inflammation. Intraocular lenses were removed in 23 cases (28%), usually to facilitate vitrectomy. Only 1 of 57 eyes with bacterial endophthalmitis could not be sterilized while the intraocular lens was retained. Visual acuity of 20/400 or better was achieved in 63% of culture positive cases, in 78% of eyes infected with S. epidermidis, and in 94% of culture negative eyes.

Ulcerative keratitis associated with contact lens wear.

From October 1982 through June 1986, 658 patients developed ulcerative keratitis. In 196 of these patients it was contact lens-related. Fifty-nine patients wore extended-wear contact lenses for cosmetic purposes. On culture, Pseudomonas species was the organism most frequently isolated from the ulcers associated with contact lens wear. No cases of fungal keratitis were found in the contact lens group as compared to 40 cases (17%) in the noncontact lens group. Compared to results of a similar study covering January 1977 through September 1982, current results showed a larger number of patients using extended-wear lenses for cosmetic reasons (59 vs one) and overall younger age.

Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study One Year After Surgery

The Prospective Evaluation of Radial Keratotomy (PERK) study is a nine-center, self-controlled clinical trial of a standardized technique of radial keratotomy in 435 patients who had physiologic myopia with a preoperative refraction between -2.00 and -8.00 diopters. The surgical technique consisted of eight incisions using a diamond micrometer knife with blade length determined by intraoperative ultrasonic pachymetry and the diameter of central clear zone determined by preoperative refraction. At one year after surgery, myopia was reduced in all eyes; 60% were within +/- 1.00 diopter of emmetropia; 30% were undercorrected and 10% were overcorrected by more than 1.00 diopter (range of refraction, -4.25 to +3.38 D). Uncorrected visual acuity was 20/40 or better in 78% of eyes. The operation was most effective in eyes with a refraction between -2.00 and -4.25 diopters. Thirteen percent of patients lost one or two Snellen lines of best corrected visual acuity. However, all but three eyes could be corrected to 20/20. Ten percent of patients increased astigmatism more than 1.00 diopter. Disabling glare was not detected with a clinical glare tester, but three patients reduced their driving at night because of glare. Between six months and one year, the refraction changed by greater than 0.50 diopters in 19% of eyes.

Late Onset Endophthalmitis Associated with Filtering Blebs

Thirty-six cases of late onset endophthalmitis in patients with filtering blebs are presented. Onset of endophthalmitis ranged from 4 months to 60 years after bleb formation. Possible contributing factors could be identified only in a minority of patients. Aqueous, vitreous or both were cultured in all cases. Eighty-three percent of eyes were culture positive. Streptococci were the most frequent causative organisms, isolated from 57% of culture positive eyes. Twenty-three percent of eyes grew Hemophilus influenzae. Only two cases were caused by staphylococci. In general, the visual outcome was poor, probably primarily due to the virulence of the infecting organisms. Endophthalmitis remains a risk even many years after creation of a filtering bleb. The microbiologic spectrum in this clinical setting is considerably different from that of recent postoperative endophthalmitis. Based on the bacteriology and clinical course of the patients presented, recommendations for management are discussed.

Microbial Endophthalmitis Resulting from Ocular Trauma

Twenty-seven cases of culture-positive endophthalmitis that developed after ocular trauma were reviewed. The intraocular culture specimens showed a virulent microbiologic spectrum with Bacillus sp as the most common isolate (8 eyes). The visual prognosis was poor, with only 22% of patients retaining 20/400 or better vision. This level of vision was achieved in 2 of 22 (9%) bacterial cases and in four of five (80%) fungal cases. Retinal detachment (5 cases) or retinal breaks (2 cases) at the time of the initial injury had a uniformly poor visual prognosis. Postoperative retinal detachment not associated with phthisis bulbi occurred in five eyes, three of which had successful retinal reattachment surgery. Delayed onset retinal detachment after successful initial management of traumatic endophthalmitis had a greater frequency of successful retinal reattachment surgery.

Exogenous Fungal Endophthalmitis

The authors report 19 cases of exogenous fungal endophthalmitis managed between 1969 and 1986. These developed after ocular surgery (5 cases), after trauma (6 cases), by intraocular spread from fungal keratitis (4 cases), after therapeutic keratoplasty for keratitis (3 cases), and by spontaneous infection of a filtration bleb (1 case). The diagnosis of endophthalmitis was made from 3 days to 4 months after surgery or trauma. In patients with preexisting keratitis, endophthalmitis was noted 2 weeks to 5 months after the onset of keratitis. Seventeen cases were caused by filamentous fungi; two were caused by yeast. Fusarium (6 cases) and Acremonium (3 cases) accounted for approximately one half of the isolates. Forty-two percent of the patients recovered 20/400 or better acuity (3 of 5 postoperative cases, 4 of 6 trauma cases, and 1 of 4 keratitis cases). The clinical and microbiologic features of these cases are presented, and recommendations for management are offered.

Results in the Treatment of Postoperative Endophthalmitis

A retrospective study of 40 cases of postoperative endophthalmitis was conducted between July 1979 and May 1981. Treatment consisted of topical, periocular, and systemic antibiotics, as well as the use of intraocular antibiotic injection in all cases. In addition, 22 cases had a diagnostic and therapeutic vitrectomy. The cases selected for vitrectomy included cases with worse presenting visual acuity, higher percentage of positive culture results, and more virulent organisms. Because of a poor clinical response to initial therapy, 13 cases had repeat intraocular cultures performed between 24 and 72 hours at the time of repeat intraocular antibiotic injection. All 13 repeat intraocular cultures were negative. Complications of the treatment included four retinal detachments (three in the vitrectomy group). In the culture-positive cases, a final visual acuity of 20/400 or better was achieved in 13 of 29 cases (45%), and complete loss of vision (NLP) resulted in 10 of the 29 cases (34%).