Open globe injuries with positive intraocular cultures

Open Globe Injuries with Positive Intraocular Cultures Factors Influencing Final Visual Acuity Outcomes Douglas F. Lieb, MD, Ingrid U. Scott, MD, MPH, Harry W. Flynn Jr, MD, Darlene Miller, MPH, William J. Feuer, MS Purpose: To investigate the clinical features influencing final visual acuity outcomes of eyes with positive intraocular cultures after open globe injuries. Design: Retrospective, consecutive, interventional case series. Participants: Thirty-seven patients. Methods: The medical records were reviewed of all patients with positive intraocular cultures after open globe injuries treated at Bascom Palmer Eye Institute between January 1, 1995, and December 31, 2001. Main Outcome Measures: Final visual acuity. Clinical features investigated included the following: (1) presence or absence of clinical endophthalmitis; (2) virulence of the cultured organism (coagulase-negative Staphylococci, Corynebacterium, and Propionibacterium acnes were classified as nonvirulent organisms, whereas all other organisms were classified as virulent organisms); (3) presence of intraocular foreign body (IOFB); (4) presence of retinal detachment; (5) interval between ocular injury and surgical repair; (6) severity of vision loss at presentation; (7) zone of injury; (8) wound length; and (9) presence of vitreous hemorrhage. Results: The study included 37 eyes of 37 patients with a mean age of 30 years (range, 18 months– 85 years) and a median follow-up of 13 months (range, 1–71 months). Study eyes were stratified into two groups: group 1 eyes (n ⫽ 16) were those in which clinical endophthalmitis did not develop, whereas group 2 eyes (n ⫽ 21) were those in which clinically diagnosed endophthalmitis developed at some point during their clinical course. Presenting visual acuity was similar in the two groups (mean logarithm of the minimum angle of resolution [logMAR] acuity, 1.91 and 2.22 [Snellen equivalents, 2/162 and 2/331] respectively; P ⫽ 0.33). Final acuities in the two groups were different, but not to a statistically significant level (mean logMAR acuity, 1.14 and 2.05 [Snellen equivalents, 20/276 and 2/224], respectively; P ⫽ 0.069). In group 1, final visual acuity ranged from 20/20 to no light perception (median acuity, 20/186); 12 eyes (75.0%) achieved a final visual acuity of 20/400 or better. In group 2, final visual acuity ranged from 20/25 to no light perception (median acuity, 7/200); of 20 eyes with known final visual acuity, 10 (50.0%) retained 20/400 or better vision. In group 1, three eyes (19%) eyes had virulent organisms, and 13 eyes (81%) had nonvirulent organisms. In group 2, 12 eyes (57%) had virulent organisms, and nine eyes (43%) had nonvirulent organisms. A final acuity of 20/60 or better was achieved in 14 eyes (41%), and a final acuity of 20/400 or better was achieved in 22 eyes (59%). Better presenting visual acuity (P ⫽ 0.038), culture of a nonvirulent organism (P ⫽ 0.011), lack of a retinal detachment (P ⫽ 0.002), absence of clinical endophthalmitis (P ⫽ 0.069), and shorter wound length (P ⫽ 0.024) were associated with better visual acuity outcome. In four of six eyes (67%) with both an IOFB and clinical endophthalmitis (group 2), the final visual acuity was no light perception (IOFB was not itself significantly associated with final visual acuity; P ⫽ 0.11). Conclusions: Among eyes with positive intraocular cultures after open globe injury, the visual prognosis is guarded. Clinical features associated with better visual acuity outcomes include better presenting visual acuity, culture of a nonvirulent organism, lack of a retinal detachment, absence of clinical endophthalmitis, and shorter wound length. Ophthalmology 2003;110:1560 –1566 © 2003 by the American Academy of Ophthalmology.

Originally received: August 13, 2002. Accepted: December 4, 2002. Manuscript no. 220561. From the Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida. Supported in part by Research to Prevent Blindness, Inc., New York, New York. Correspondence to Ingrid U. Scott, MD, MPH, Bascom Palmer Eye Institute, P.O. Box 016880, Miami, FL 33101. E-mail: iscott@bpei. med.miami.edu.

1560

© 2003 by the American Academy of Ophthalmology Published by Elsevier Inc.

Endophthalmitis is a potentially devastating complication of open globe injury.1–11 Posttraumatic endophthalmitis has been reported to occur in approximately 4% to 8% of eyes with open globe injury1,7,8 and in up to 30% of eyes with open globe injuries sustained in rural settings.10 In a series of 496 patients with retained intraocular foreign bodies (IOFBs), endophthalmitis developed in 6.9%.5 Generally accepted risk factors for posttraumatic endophthalmitis include delayed primary closure of open globe injuries, presISSN 0161-6420/03/$–see front matter doi:10.1016/S0161-6420(03)00497-4

Lieb et al 䡠 Open Globe Injuries and Endophthalmitis Table 1. Eyes with Positive Cultures and No Clinically Diagnosed Endophthalmitis (Group 1) Visual Acuity Patient No. Initial* Final 1 2 3 4 5 6

HM 20/40 HM 20/40 6/200 NLP F/F 6/100 CF 2/200 20/60

20/25

7

3/200 20/20

8

2/200

9

HM

20/40

10 11

20/40 20/50

20/400 20/30

12 13

LP HM

20/400 20/20

14 15 16

HM NLP 1/200 20/20 HM 20/400

2/200

Initial Pars Later Pars Intraocular Initial Plana Plana Foreign Antibiotics Vitrectomy Vitrectomy Body Sensitivities

Organism Isolated

Age

Gender

Eye

Staphylococcus epidermidis Staphylococcus epidermidis Corynebacterium Staphylococcus epidermidis Bacillus species Staphylococcus epidermidis Coagulase-negative Staphylococcus Corynebacterium Corynebacterium Trichosporium Dematiaceous fungi Escherichia coli Staphylococcus hominis Staphylococcus aureus Propionibacterium acnes Staphylococcus epidermidis Staphylococcus epidermidis Staphylococcus hemolyticus Staphylococcus epidermidis Propionibacterium acnes Staphylococcus epidermidis Staphylococcus homini Staphylococcus sciuri Staphylococcus epidermidis Staphylococcus warneri Propionibacterium acnes

8 yrs 32 yrs 24 yrs 18 mos 36 yrs

M M M M M

Right Right Right Left Left

N N Y N N

N Y Y Y Y

Y N Y Y N

N Y Y N Y

Y Y Y NA NA

22 yrs

M

Left

N

Y

N

Y

NA

36 yrs

M

Left

N

N

N

Y

NA

85 yrs

F

Left

N

N

N

N

NA

25 yrs

M

Left

N

Y

N

Y

NA

34 yrs 24 yrs

M M

Right Right

Y Y

Y Y

N Y

Y Y

NA Y

12 yrs 47 yrs

M M

Right Right

N N

N N

Y N

N N

NA NA

31 yrs 31 yrs 39 yrs

M M M

Right Right Right

N Y Y

Y N Y

N N N

Y N Y

NA Y Y

F ⫽ female; F/F ⫽ fix and follow; HM ⫽ hand motions; LP ⫽ light perception; M ⫽ male; N ⫽ no; NA ⫽ not applicable; NLP ⫽ no light perception; U ⫽ unable to culture organism and sensitivities unknown; Y ⫽ yes. No eyes received intravitreal antibiotics after the initial open globe repair. *At the time of initial open globe repair.

ence of an IOFB, and rural setting of injury or soil-associated injury. Other reported risk factors include rupture of the crystalline lens and age more than 50 years. Gram-positive organisms such as Bacillus, Staphylococci, and Streptococci are frequently isolated pathogens.1,4,5,9 –11 The current study describes the microbiologic and treatment outcomes of eyes with positive intraocular cultures after open globe injuries and investigates clinical features associated with visual acuity outcomes.

Materials and Methods The study protocol was approved by the University of Miami School of Medicine Institutional Review Board. The medical records were reviewed of all patients with positive intraocular cultures after open globe injuries treated at Bascom Palmer Eye Institute between January 1, 1995, and December 31, 2001. Thirtynine eyes of 39 patients were identified. One patient, an 84-yearold man, died 2 days after his open globe injury and he was excluded from the study because of inadequate follow-up. Another patient, a 10-year-old with a penetrating ocular injury resulting from a fish hook, had scant growth of a gram-negative filamentous bacterium that could not be speciated. This patient had 20/20 visual acuity at presentation and at the final follow-up examination, but was excluded because no specific organism was identified. In the current study, the criteria used for making a clinical diagnosis of endophthalmitis included one or more of the follow-

ing clinical features: (1) marked intraocular inflammation with or without hypopyon; (2) retinal periphlebitis;12,13 and (3) marked exudate around an intraocular foreign body. Presenting and final visual acuities, organism cultured, management, presence of an IOFB, and antibiotic sensitivities of cultured organisms were recorded for each patient. For the purposes of analysis, eyes were divided into two groups: group 1 included eyes in which clinically diagnosed endophthalmitis did not develop; group 2 included eyes in which clinically diagnosed endophthalmitis developed at some point during their clinical course. Eyes were divided into those with virulent organisms and those with nonvirulent organisms. Coagulase-negative Staphylococci, Corynebacterium, and Propionibacterium acnes were considered to be nonvirulent. Other cultured organisms were considered to be virulent. Eyes with polymicrobial growth were considered to harbor virulent organisms if one or more of the cultured organisms was considered virulent. The following potential risk factors were evaluated for their influence on final visual acuity: presenting visual acuity, absence or presence of clinical endophthalmitis (groups 1and 2, respectively); virulence of the cultured organism; presence of an IOFB; presence of retinal detachment at any time during management; interval between ocular injury and surgical repair; presenting acuity; zone of injury; wound length; and presence of vitreous hemorrhage at the time of wound closure. The influence of risk factors on final logarithm of the minimum angle of resolution (logMAR) visual acuity was examined with t tests for dichotomous variables, one-way analysis of variance for polychotomous variables, and with Pearson correlation coefficients for continuous variables. Visions of count fingers, hand motions, light perception, and no

1561

Ophthalmology Volume 110, Number 8, August 2003 Table 2. Eyes with Positive Cultures and Clinically Diagnosed Endophthalmitis (Group 2) Visual Acuity

Initial Pars Later Pars Intraocular Plana Foreign Initial Plana Body Sensitivities Gender Eye Antibiotics* Vitrectomy† Vitrectomy‡

Patient No.

Initial

Final

Organism Isolated

Age

1

NA

20/40

19 mos

M

Left

Y

Y

N

N

Y

2 3 4 5 6 7 8

LP NLP HM 20/400 HM NA LP

20/60 NLP 20/60 NLP NLP NLP NLP

74 44 66 25 82 2 9

yrs yrs yrs yrs yrs yrs yrs

F M M M F M M

Left Left Right Right Right Right Left

Y N N Y N§ Y Y

N N Y Y N N Y

N N N N Y N N

N Y N Y N N Y

Y NA NA Y U Y Y

9 10 11 12 13

HM 20/400 CF 20/300 NA

HM 20/50 20/300 20/50 NA

30 13 6 21 2

yrs yrs yrs yrs yrs

M M M M F

Left Left Right Right Left

Y Y N§ N§ N§

N N N N N

Y N Y Y Y

N N N N N

Y Y Y Y Y

14 15

CF HM

20/400 20/40

30 yrs 16 yrs

M M

Right Right

N§ N§

N N

Y Y

N N

Y Y

16 17 18 19

20/60 HM LP LP

20/25 NLP LP 20/400

39 26 48 46

yrs yrs yrs yrs

M M M M

Right Right Right Right

N§ Y Y Y

N Y Y Y

Y Y Y Y

N Y Y Y

Y Y Y

20 21

CF 20/200

LP 3/200

Streptococcus salivarius Streptococcus oralis Neiseria mucosa Staphylococcus epidermidis Bacillus cereus Staphylococcus epidermidis Bacillus cereus fungus Streptococcus viridans Bacillus cereus Enterococcus gallinrum Staphyllococcus aureus Staphylococcus epidermidis Staphylococcus epidermidis Bacillus cereus Staphylococcus epidermidis Coagulase-negative staphylococcus Staphylococcus epidermidis Streptococcus sanguis Coagulase-negative staphylococcus Coagulase-negative staphylococcus Moraxella Bacillus cereus Bacillus species Enterococcus gallinurum Enterobacter clocae Escherichia coli Leclercia adecarboxylata Curvularia species Propionibacterium acnes Pseudomonas species Aureobacterium species Sporothrix schenckii Cryptococcus laurentii

10 yrs 29 yrs

F M

Right Left

N N§

N N

Y Y

N N

Y

㛳

㛳

CF ⫽ counting fingers; F ⫽ female; HM ⫽ hand motion; LP ⫽ light perception; M ⫽ male; N ⫽ no; NA ⫽ not applicable; NLP ⫽ no light perception; U ⫽ sensitivities unavailable; Y ⫽ yes. *Antibiotics at the time of initial open globe repair. † At the time of initial open globe repair. ‡ After initial open globe repair. § No initial antibiotics, but subsequent antibiotics given. 㛳 All bacteria sensitive, fungus resistant.

light perception were assigned values in a fashion similar to that reported previously.14 Additionally, the association of risk factors was examined with final acuity split into clinically meaningful categories (ⱖ20/60 vs. ⬍20/60 and ⱖ20/400 vs. ⬍20/400). Proportions were compared with Fisher exact test and the chi-square test.

Results The study included 16 patients in group 1 and 21 patients in group 2 (Tables 1, 2, 3, and 4). The mean age of the patients (31 and 30 years, respectively; P ⫽ 0.89), mean logMAR presenting visual acuity (1.91 and 2.22 [Snellen equivalents, 2/162 and 2/331], respectively; P ⫽ 0.33), and median duration of follow-up (14.3 and 9.3 months, respectively; P ⫽ 0.037) were similar between the two groups. In group 1, one patient was female. In group 2, four patients were female. An IOFB was present in 10 of 16 eyes (62.5%) in group 1 compared with 6 of 21 eyes (28.5%) in group

1562

2 (P ⫽ 0.052). Intraocular antibiotics were injected at the time of open globe repair in five eyes (31.3%) in group 1 and in 10 eyes (47.6%) in group 2; intraocular antibiotics were injected subsequently in 0 eyes (0%) and nine eyes (42.9%) in the two groups, respectively (P ⫽ 0.005). Ten eyes (62.5%) in group 1 underwent pars plana vitrectomy on presentation compared with 7 eyes (33.3%) in group 2 (P ⫽ 0.10); five eyes (31.3%) and 13 eyes (61.9%), respectively, subsequently underwent pars plana vitrectomy (P ⫽ 0.10). Overall, 15 of 16 eyes (94%) in group 1 and 20 of 21 eyes (95%) in group 2 underwent vitrectomy at some point during the clinical course. Among the 16 eyes in group 1 (Table 1), presenting visual acuity ranged from 20/40 to light perception (mean logMAR acuity, 1.91 [Snellen equivalent, 2/162]). Presenting visual acuity was not quantifiable more specifically than response to light in a single patient because of young age at presentation (18 months). Final visual acuity ranged from 20/20 to no light perception (median acuity, 20/186). Twelve eyes (75.0%) achieved a final visual acuity

Lieb et al 䡠 Open Globe Injuries and Endophthalmitis Table 3. Clinical Description and Time to Presentation (Group 1) Patient No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Clinical Description 7-mm corneal laceration with plant matter 3-mm corneal laceration with metallic IOFB resting on retina 3.5-mm corneal laceration with metallic IOFB resting on retina 9-mm scleral laceration resulting from glass fragment 4-mm corneal laceration with metallic IOFB embedded in retina 3-mm corneal laceration with metal IOFB resting on retina 2-mm pars plicata wound; fishhook extending into vitreous cavity Corneal graft missing after blunt injury 4-mm corneal laceration resulting from nail gun injury; posterior exit site 4-mm corneal laceration with metal IOFB resting on macula 2-mm corneal laceration with metal IOFB resting on retina 12-mm corneoscleral laceration resulting from bomb blast 8-mm corneoscleral laceration sustained while cleaning boat 10-mm corneoscleral laceration with nail gun 1.5-mm corneal laceration resulting from metal object 4-mm corneal laceration with metallic IOFB in anterior chamber

Table 4. Clinical Description and Time to Presentation (Group 2)

Time from Injury to Presentation

Patient No.

5 hrs 2 hrs

1 2

6 hrs

3

2 hrs

4

2 hrs

5

4 hrs

6

2 hrs

7

3 hrs 4 hrs

8 9

4 hrs 4 hrs 4 days

10 11 12 13

6 hrs 14 4 hrs 15 4 hrs 16 5 hrs

IOFB ⫽ intraocular foreign body.

17 18 19

of 20/400 or better. Two eyes (13%) in group 1 were no light perception at last follow-up. Among the 21 eyes in group 2 (Table 2), presenting visual acuity ranged from 20/60 to no light perception (logMAR mean acuity, 2.22 [Snellen equivalent, 2/331]). Presenting visual acuity was unknown in three patients because of very young age at the time of injury (19 –24 months). Final visual acuity ranged from 20/25 to no light perception (median acuity, 7/200). Of 20 eyes with known final visual acuity, 10 (50.0%) retained 20/400 or better vision. Final visual acuity was not given in one patient because of young age at last follow-up. Six eyes (30.0%) were no light perception at last follow-up. Overall, of the 24 (65.8%) eyes that received intravitreal antibiotics, 21 (87.5%) received antibiotics to which all of the organisms were susceptible, including 6 eyes (100%) in group 1 and 16 eyes (84.2%) in group 2 (P ⫽ 0.023). In 1 of 24 eyes receiving intravitreal antibiotics, the sensitivities of the cultured organisms were not available. In an additional two eyes, all bacterial isolates were sensitive to the injected antibiotics, but the fungal isolates were not sensitive. Final visual acuity outcomes for both groups are shown in Tables 1 and 2. Overall, a final visual acuity of 20/60 or better was achieved in 15 eyes (41%), and 14 (38%) had a final visual acuity worse than 20/400. Tables 5 and 6 display the associations between risk factors and final visual acuity. Variables associated with a better final visual acuity include the following: better presenting visual acuity (P ⫽ 0.038), presence of a nonvirulent organism (P ⫽ 0.011), lack of a retinal detachment (P ⫽ 0.002), and absence of clinical endophthalmitis (P ⫽ 0.069).

20 21

Clinical Description

Time from Injury to Presentation

4-mm corneal laceration resulting from pen 1 day 10-mm dehiscence of corneal graft resulting 1 day from blunt injury 4-mm corneoscleral laceration with 1 day retained metal IOFB 7-mm corneal laceration with retained 1 day metal IOFB 2-mm corneal laceration; metal IOFB in 2 hrs optic disc 9-mm dehiscence of limbal cataract 4 hrs incision; eye struck with tree branch Unrecognized stellate corneal laceration, Approximately 3 cause unknown days 10-mm corneoscleral laceration with 12 hrs retained vitreous debris from firecracker 5-mm corneal laceration; giant retinal tear 3 days noted 2.5-mm corneal laceration with ink pen 2 days 10-mm corneoscleral laceration resulting 3 hrs from rock thrown at school 3-mm corneoscleral laceration 2 hrs 7-mm corneal laceration resulting from 2 hrs vegetable matter 4-mm corneoscleral laceration resulting 1 hr from nail gun 2-mm corneoscleral laceration resulting 3 hrs from dart 1-mm wound at limbus resulting from 1 day sewing needle 3-mm corneoscleral laceration resulting 3 days from glass fragment 6-mm corneoscleral laceration resulting 1 hr from nail gun injury; posterior scleral rupture noted 12-mm corneoscleral laceration with 2 hrs retained intraocular sea shell; injury sustained while mowing lawn 10-mm corneoscleral laceration 1 day 2-mm corneal laceration 4 hrs

IOFB ⫽ intraocular foreign body.

At some point during their clinical course, retinal detachment developed in seven patients, including two eyes in group 1 (patients 3 and 14) and five eyes in group 2 (patients 7, 8, 18, 19, and 21). Of these seven eyes, three were enucleated or eviscerated during the follow-up period, one was no light perception at last follow-up, and only one patient achieved a final vision of 20/400 or better (20/400). The retina was attached at last follow-up in all four eyes not enucleated or eviscerated. Although presence of an IOFB was not a statistically significant prognostic factor, retained IOFB in combination with posttraumatic, clinically diagnosed endophthalmitis was associated with a particularly poor visual outcome. Of six eyes with an IOFB in group 2 (Table 2), four eyes (66.7%) had a final visual acuity of no light perception. Five of these six eyes were treated with both pars plana vitrectomy and intravitreal antibiotics at the time of presentation; one of six eyes was treated with primary evisceration because of no light perception vision on presentation (group 2, patient 3). Five of six eyes with an IOFB had positive cultures for Bacillus species, and the other eye had Moraxella (group 2, patient 17). The cultured organisms in five of six cases were sensitive to the injected intravitreal antibiotics. All bacterial organisms were

1563

Ophthalmology Volume 110, Number 8, August 2003 Table 5. Categorical Risk Factors and Their Association with Logarithm of the Minimum Angle of Resolution Final Visual Acuity

Gender, P ⫽ 0.27 Male Female Ethnicity, P ⫽ 0.33 White Black Hispanic Other Presence of clinical endophthalmitis, P ⫽ 0.069 No Yes Virulent organism, P ⫽ 0.011 No Yes Intraocular foreign body, P ⫽ 0.11 No Yes Zone of injury, P ⫽ 0.70 Cornea and limbus Up to 5 mm posterior to limbus Over 5 mm posterior to limbus Retinal detachment diagnosed during management, P ⫽ 0.002 No Yes Wound size group, P ⫽ 0.31 ⬍5 mm ⬎5 mm Presentation on day of the injury, P ⫽ 0.57 No Yes

n

Mean ⴞ Standard Deviation

[Snellen Equivalent of Mean Acuity]

32 5

1.55 ⫾ 1.51 2.36 ⫾ 1.30

[5/177] [1/229]

17 6 10 2

1.94 ⫾ 1.57 1.81 ⫾ 1.82 0.96 ⫾ 1.08 0.76 ⫾ 0.65

[5/435] [5/323] [20/182] [20/115]

16 21

1.14 ⫾ 1.32 2.05 ⫾ 1.54

[20/276] [2/224]

22 15

1.15 ⫾ 1.26 2.39 ⫾ 1.55

[20/283] [2/258]

21 16

1.31 ⫾ 1.26 2.11 ⫾ 1.69

[20/408] [2/258]

26 8

1.71 ⫾ 1.53 1.73 ⫾ 1.70

[5/256] [5/269]

2

0.76 ⫾ 0.65

[20/115]

30 7

1.50 ⫾ 1.36 1.86 ⫾ 1.68

[20/400] [LP]

21 14

1.37 ⫾ 1.49 1.90 ⫾ 1.46

[5/117] [5/397]

26 11

1.56 ⫾ 1.45 1.88 ⫾ 1.66

[5/182] [5/379]

Pathologic Feature

Group 1

Group 2

Retinal detachment Macular scar Hypotony (IOP ⬍6 mmHg) Optic atrophy Corneal scar Amblyopia history Glaucoma (IOP ⱖ30 mmHg) Epiretinal membrane

2 2 2 0 5 1 0 1 n ⫽ 16

5 1 1 1 9 3 1 0 n ⫽ 21

IOP ⫽ intraocular pressure.

detachment (P ⫽ 0.028 and P ⫽ 0.011, respectively). Shorter wound size was associated significantly with a final acuity of ⱖ20/60 (P ⫽ 0.024). Presence of vitreous hemorrhage at the time of wound closure was not a significant prognostic factor.

Discussion

sensitive to the injected antibiotics in the six eyes with an IOFB, but the fungal (Curvularia) isolate was resistant in one of six eyes (group 2, patient 19). The specific factors limiting final visual acuity are identified in Table 7. One patient in group 2 had optic atrophy resulting from an impact from a foreign body. The association between each risk factor and final visual acuity grouped into clinically meaningful categories is displayed in Table 8. In these analyses, a nonvirulent organism was significantly associated with both a final visual acuity of ⱖ20/60 (P ⫽ 0.047) and a final acuity of ⱖ20/400 (P ⫽ 0.016), as was lack of a retinal Table 6. Pearson Correlations (r) of Continuous Risk Factors with Logarithm of the Minimum Angle of Resolution Final Visual Acuity Risk Factor

r

P Value

Presenting acuity Months followed up Wound length Days from injury to presentation Age at surgery

0.35 0.23 0.24 0.20 0.02

0.038 0.17 0.16 0.24 0.90

1564

Table 7. Factors Contributing to Reduced Final Visual Acuity

The visual acuity outcomes in the current study are similar to those found in other reports of posttraumatic endophthalmitis (Table 9). 1,5,8 –11,15,16 A final visual acuity of 20/400 or better was achieved in 22 patients (61%) overall and in 10 patients (50.0%) with culture-positive, clinically suspected endophthalmitis. Previous endophthalmitis studies have reported the trend to better visual outcomes when the infection was caused by less virulent organisms.1,5,11 However, useful vision may be retained in some eyes with posttraumatic endophthalmitis resulting from Bacillus species.2 In a review of 492 eyes with IOFBs in the National Eye Trauma Survey registry, 21 of 26 eyes (81%) with culture-proven posttraumatic endophthalmitis retained 20/ 200 or better vision.8 In the current study, no light perception vision was present at final follow-up in four of six eyes with both clinical endophthalmitis and an IOFB. Positive intraocular cultures at the time of surgery have been reported with primary pars plana vitrectomy for indications other than open globe injury,17 extracapsular cataract extraction,18 and uncomplicated phacoemulsification procedures19,20 (Table 10). It is possible that some of the study eyes in group 1 had positive intraocular cultures associated with standard pars plana vitrectomy not associated with the open globe injury, because 9 of 16 eyes in group 1 were treated with initial vitrectomy. However, in the current series, 7 of 16 isolates were polymicrobial, whereas in the series of patients with positive intraocular cultures associated with primary pars plana vitrectomy reported by Cohen et al17 the two patients with positive intraocular cultures included only one organism. Also, 10 of 16 study eyes in group 1 had an IOFB, a reported risk factor for posttraumatic endophthalmitis.7 Seven patients in the current series developed retinal detachment sometime during their clinical course, including four eyes with retinal detachment on presentation. Options for the management of concurrent endophthalmitis and retinal detachment include: (1) use of antibiotics in the irrigating fluid; (2) injection of a reduced dose of intravitreal

Lieb et al 䡠 Open Globe Injuries and Endophthalmitis Table 8. Patient Risk Factors and Their Association with Final Visual Acuity Grouped into Clinically Meaningful Categories Categorical Variables at Presentation, n (%)

Final Acuity >20/60

Gender Male Female Ethnicity White Black Hispanic Other Presenting acuity ⱖ20/400 ⬍20/400–HM LP, NLP Clinical endophthalmitis No Yes Virulent organism No Yes IOFB Yes No Zone of injury Cornea and limbus ⱕ5 mm posterior ⬎5 mm posterior Wound size ⱖ5 mm Yes No Vitreous hemorrhage (at time of surgery) Yes No Presented same day as injury [N (%)] Yes No Retinal detachment at any time during management Yes No

Final Acuity >20/400

P Value* 0.63

P Value* 0.14

14 (44%) 1 (20%)

21 (66%) 1 (20%) 0.45

0.31

1 (50%) 3 (50%) 5 (29%) 6 (60%)

2 (100%) 3 (50%) 9 (53%) 8 (80%) 0.22

0.46

6 (67%) 6 (35%) 2 (29%)

7 (78%) 9 (53%) 4 (57%) 0.34

0.18

8 (50%) 7 (33%)

12 (75%) 10 (48%) 0.047

0.016

12 (55%) 3 (20%)

17 (77%) 5 (33%) 0.50

0.34

10 (48%) 5 (31%)

14 (67%) 8 (50%) 0.94

0.43

11 (42%) 3 (38%) 1 (50%)

14 (54%) 5 (63%) 2 (100%) 0.30

0.72

11 (52%) 4 (29%)

14 (67%) 8 (57%) 1.0

1.0

4 (36%) 11 (42%)

7 (64%) 15 (58%) 0.73

0.73

10 (39%) 5 (46%)

16 (62%) 6 (55%) 0.028

0.011

0 (0%) 15 (50%)

1 (14%) 21 (70%)

Continuous Variables at Presentation, Mean (SD)

Final Acuity >20/60

Final Acuity <20/60

P Value

Age Months followed up Wound length

30.7 (20.0) 18.5 (18.4) 4.0 (2.7)

30.0 (22.8) 20.8 (20.7) 6.5 (3.4)

0.11 0.73 0.024

†

Final Acuity >20/400

Final Acuity <20/400

P Value†

28.6 (19.1) 16.4 (17.2) 5.2 (3.5)

32.6 (24.9) 24.9 (22.2) 5.8 (3.1)

0.59 0.20 0.63

HM ⫽ hand motion; IOFB ⫽ intraocular foreign body; LP ⫽ light perception; NLP ⫽ no light perception. *P values by Fisher exact test for dichotomous variables and chi-square for polychotomous variables. † P values by two-sample two-tailed t test.

antibiotics into a gas-filled eye; and (3) injection of the usual dose of intravitreal antibiotics into a 50% gas-filled eye.21 In the current series, concurrent retinal detachment and endophthalmitis were managed by injection of the usual dose of intravitreal antibiotics into a 50% gas-filled eye. The decision of whether to administer prophylactic intraocular antibiotics in patients with open globe injuries and no hypopyon or other classic signs of endophthalmitis remains controversial. In the opinion of the authors, clinical factors that warrant consideration of prophylactic intraocu-

lar antibiotics include retinal periphlebitis, purulence around a foreign body, soil-related injury, and delayed primary wound closure. The antibiotics the authors generally prefer are vancomycin for coverage of gram-positive organisms (including Bacillus) and ceftazidime for coverage of gram-negative organisms. In the current series, a positive intraocular culture in the setting of open globe injury is generally associated with a poor visual outcome; however, 22 of these patients (59.5%) achieved a visual acuity of 20/400 or better. Clinical fea-

1565

Ophthalmology Volume 110, Number 8, August 2003 Table 9. Reported Visual Acuity Outcomes after Management of Posttraumatic Endophthalmitis Number Year of Eyes

Authors

Brinton et al* 1984 Affeldt et al† 1987 Williams et al‡ 1988 Boldt et al§ 1989 㛳 Thompson et al 1993 Alfaro et al¶ 1994 Verbraeken and Rysselaeveⱅ 1994 Thompson et al** 1995 Current study 2002

19 27 14 24 34 36 25 41 20††

Final Vision >20/400

Intraocular Foreign Body

42.1% 22.0% (or better) 64.3% 29.2% 58.0% 27.8% 16.0% 27.0% (20/50) 50.0%

57.9% 29.6% 100% 54% 100% 16.7% 52% 39.0% 28.6%

*Brinton GS, Topping TM, Hyndiuk RA, et al. Posttraumatic endophthalmitis. Arch Ophthalmol 1984;192:547–50. † Affeldt JC, Flynn HW Jr, Forster RK, et al. Microbial endophthalmitis resulting from ocular trauma. Ophthalmology 1987;94:407–13. ‡ Williams DF, Mieler WF, Abrams GW, Lewis H. Results and prognostic factors in penetrating ocular injuries with retained intraocular foreign bodies. Ophthalmology 1988;95:911– 6. § Boldt HC, Pulido JS, Blodi CF, et al. Rural endophthalmitis. Ophthalmology 1989;96:1722– 6. 㛳 Thompson JT, Parver LM, Enger CL, et al. Infectious endophthalmitis after penetrating injuries with retained intraocular foreign bodies. National Eye Trauma System. Ophthalmology 1993;100:1468 –74. ¶ Alfaro DV, Roth D, Liggett PE. Posttraumatic endophthalmitis. Causative organisms, treatment, and prevention. Retina 1994;14:206 –11. ⱅ Verbraeken H. Rysselaere M. Post-traumatic endophthalmitis. Eur J Ophthalmol 1994;4:1–5. **Thompson WS, Rubsamen PE, Flynn HW Jr, et al. Endophthalmitis after penetrating trauma. Risk factors and visual acuity outcomes. Ophthalmology 1995;102:1696 –701. †† Snellen visual acuity could not be assessed in one infant in group 2.

tures significantly associated with a better final visual acuity include a better presenting visual acuity, culture of a nonvirulent organism, lack of retinal detachment, absence of clinical endophthalmitis, and shorter wound length. Table 10. Positive Intraocular Cultures during Elective Surgery

Authors Cohen et al* Samad et al† Gimbel et al‡ Sherwood et al§

No. (%) of Total No. eyes with Year of Eyes (ⴙ) culture 1992 1995 1996 1989

33 103 97 100

2 (6) 5 (5) 1 (1) 29 (29)

Procedure PPV Phacoemulsification Phacoemulsification ECCE

ECCE ⫽ extracapsular cataract extraction; PPV ⫽ pars plana vitrectomy; (⫹) ⫽ positive. *Cohen SM, Benner JD, Landers MB III, Morse LS. Intraocular fluid cultures after primary pars plana vitrectomy. Am J Ophthalmol 1992;114: 697–9. † Samad A, Solomon LD, Miller MA, Mendelson J. Anterior chamber contamination after uncomplicated phacoemulsification and intraocular lens implantation. Am J Ophthalmol 1995;120:143–50. ‡ Gimbel HV, Sun R, DeBroff BM, Yang HM. Anterior chamber fluid cultures following phacoemulsification and posterior chamber lens implantation. Ophthalmic Surg Lasers 1996;27:121– 6. § Sherwood DR, Rich WJ, Jacob JS, et al. Bacterial contamination of intraocular and extraocular fluids during extracapsular cataract surgery. Eye 1989;3:308 –12.

1566

References 1. Thompson WS, Rubsamen PE, Flynn HW Jr, et al. Endophthalmitis after penetrating trauma. Risk factors and visual acuity outcomes. Ophthalmology 1995;102:1696 –701. 2. Foster RE, Martinez JA, Murray TG, et al. Useful visual outcomes after treatment of Bacillus cereus endophthalmitis. Ophthalmology 1996;103:390 –7. 3. Mieler WF, Ellis MK, Williams DF, Han DP. Retained intraocular foreign bodies and endophthalmitis. Ophthalmology 1990;97:1532– 8. 4. Alfaro DV, Roth DB, Laughlin RM, et al. Paediatric posttraumatic endophthalmitis. Br J Ophthalmol 1995;79:888 –91. 5. Verbraeken H, Rysselaere M. Post-traumatic endophthalmitis. Eur J Ophthalmol 1994;4:1–5. 6. Mittra RA, Mieler WF. Controversies in the management of open-globe injuries involving the posterior segment. Surv Ophthalmol 1999;44:215–25. 7. Duch-Samper AM, Menezo JL, Hurtado-Sarrio M. Endophthalmitis following penetrating eye injuries. Acta Ophthalmol Scand 1997;75:104 – 6. 8. Thompson JT, Parver LM, Enger CL, et al. Infectious endophthalmitis after penetrating injuries with retained intraocular foreign bodies. National Eye Trauma System. Ophthalmology 1993;100:1468 –74. 9. Alfaro DV, Roth D, Liggett PE. Posttraumatic endophthalmitis. Causative organisms, treatment, and prevention. Retina 1994;14:206 –11. 10. Boldt HC, Pulido JS, Blodi CF, et al. Rural endophthalmitis. Ophthalmology 1989;96:1722– 6. 11. Affeldt JC, Flynn HW Jr, Forster RK, et al. Microbial endophthalmitis resulting from ocular trauma. Ophthalmology 1987; 94:407–13. 12. Jeng BH, Kaiser PK, Lowder CY. Retinal vasculitis and posterior pole “hypopyons” as early signs of acute bacterial endophthalmitis. Am J Ophthalmol 2001;131:800 –2. 13. Packer AJ, Weingeist TA, Abrams GW. Retinal periphlebitis as an early sign of bacterial endophthalmitis. Am J Ophthalmol 1983;96:66 –71. 14. Scott IU, Schein OD, West S, et al. Functional status and quality of life measurement among ophthalmic patients. Arch Ophthalmol 1994;112:329 –35. 15. Brinton GS, Topping TM, Hyndiuk RA, et al. Posttraumatic endophthalmitis. Arch Ophthalmol 1984;192:547–50. 16. Williams DF, Mieler WF, Abrams GW, Lewis H. Results and prognostic factors in penetrating ocular injuries with retained intraocular foreign bodies. Ophthalmology 1988;95:911– 6. 17. Cohen SM, Benner JD, Landers MB III, Morse LS. Intraocular fluid cultures after primary pars plana vitrectomy. Am J Ophthalmol 1992;114:697–9. 18. Sherwood DR, Rich WJ, Jacob JS, et al. Bacterial contamination of intraocular and extraocular fluids during extracapsular cataract surgery. Eye 1989;3:308 –12. 19. Samad A, Solomon LD, Miller MA, Mendelson J. Anterior chamber contamination after uncomplicated phacoemulsification and intraocular lens implantation. Am J Ophthalmol 1995;120:143–50. 20. Gimbel HV, Sun R, DeBroff BM, Yang HM. Anterior chamber fluid cultures following phacoemulsification and posterior chamber lens implantation. Ophthalmic Surg Lasers 1996;27: 121– 6. 21. Foster RE, Rubsamen PE, Joondeph BC, et al. Concurrent endophthalmitis and retinal detachment. Ophthalmology 1994;101:490 – 8.