Prospective randomized comparison of 3-day versus 1-hour preoperative ofloxacin prophylaxis for cataract surgery

Prospective Randomized Comparison of 3Day versus 1-Hour Preoperative Ofloxacin Prophylaxis for Cataract Surgery Christopher N. Ta, MD,1 Peter R. Egbert, MD,1 Kuldev Singh, MD,1 Erin M. Shriver, MD,1 Mark S. Blumenkranz, MD,1 Herminia Min˜o de Kaspar, PhD1,2 Purpose: To determine the efficacy of reducing conjunctival bacterial flora with topical ofloxacin when given for 3 days compared with 1 hour before surgery. Design: Prospective, randomized, controlled trial. Participants: Ninety-two eyes from 89 patients were randomized to a control group (48 eyes) or study group (44 eyes). Methods: All patients from both groups received topical ofloxacin 0.3% 1 hour before surgery and a 5% povidone iodine scrub of the periorbital area before surgery. The patients in the study group received additional ofloxacin four times daily for 3 days before surgery. Main Outcome Measures: Conjunctival cultures were obtained at five separate time points and were inoculated in solid and liquid culture media. The presence of bacteria was determined, quantified, and identified. Results: Forty-two percent of eyes in the control group had positive conjunctival culture immediately before surgery, compared with 19% of eyes in the study group (P ⬍ 0.05). Immediately after surgery, 34% and 14% of eyes had positive cultures in the control and study groups, respectively (P ⬍ 0.05). Quantitatively, fewer bacteria were isolated from eyes in the study group compared with those in the control group for culture samples that were obtained both before povidone iodine scrub and at the conclusion of surgery (P ⱕ 0.05). Conclusions: The application of topical ofloxacin for 3 days before surgery appears to be more effective in eliminating bacteria from the conjunctiva than an application of ofloxacin 1 hour before surgery. Ophthalmology 2002;109:2036 –2041 © 2002 by the American Academy of Ophthalmology, Inc.

Postoperative endophthalmitis is a rare but potentially devastating complication of intraocular surgery. The incidence of endophthalmitis after cataract surgery is approximately 0.082%.1 It is believed that bacteria most commonly originate from the eyelids and conjunctiva in cases of endophthalmitis.2 It is hypothesized that decreasing the number of bacteria on the eyelids and conjunctiva reduces the risk of developing endophthalmitis after surgery. Povidone (polyvinylpyrrolidone) iodine has been shown to be an effective antiseptic for use in ophthalmic surOriginally received: November 12, 2001. Accepted: March 22, 2002. Manuscript no. 210610. 1 Department of Ophthalmology, School of Medicine, Stanford University, Stanford, California. 2 Department of Ophthalmology, Ludwig-Maximilians-University, Munich, Germany. Presented at the annual meeting of the American Academy of Ophthalmology, New Orleans, Louisiana, November 2001. Supported by the Edward E. Hills Fund, San Francisco, California; Allergan, Inc., Irvine, California; and Hannelore-Georg Zimmerman Foundation, Munich, Germany. The authors have no proprietary interest in any of the products used in this study. Reprint requests to Christopher N. Ta, MD, Department of Ophthalmology, Stanford University, 900 Blake Wilbur, 3rd Floor, Stanford, CA 94304. E-mail: [email protected].

2036

© 2002 by the American Academy of Ophthalmology, Inc. Published by Elsevier Science Inc.

gery.3–9 The application of a topical antibiotic in addition to povidone iodine further reduces the number of bacteria recovered from the conjunctiva.5 The choice of antibiotic used for endophthalmitis prophylaxis is important. Broadspectrum bactericidal agents, such as fluoroquinolones, may be ideal. The optimum perioperative antibiotic regimen for endophthalmitis prophylaxis has not yet been determined. Some ophthalmologists recommend using a topical antibiotic for several days before surgery, whereas others believe that the use of antibiotic on the day of surgery is adequate.10 The goal of this study is to compare the effectiveness of a topical antibiotic in reducing or eliminating conjunctival bacterial flora when given only 1 hour before, versus for 3 days before, cataract surgery.

Materials and Methods A total of 92 eyes from 89 patients participated in this randomized, prospective study. After approval by the Institutional Review Board at Stanford University, appropriate informed consent for the study was obtained from each patient undergoing cataract surgery. None of the study participants had been diagnosed as having an active ocular infection before surgery. Patients taking topical and systemic medications were not excluded from the study except for those taking topical ocular antibiotics. The Excel software program ISSN 0161-6420/02/$–see front matter PII S0161-6420(02)01236-8

Ta et al 䡠 Prophylactic Antibiotic for Cataract Surgery

Figure 1. Percent of positive conjunctival culture from thioglycolate broth. Number in bar graph indicates number of eyes cultured. *P ⬍ 0.05.

(Microsoft Inc., Redmond, Washington) was used to generate random numbers that were assigned to each patient. Patients were randomized to either a control group (48 eyes) or study group (44 eyes). All eyes received one drop of topical ofloxacin 0.3% every 5 minutes, for a total of three applications, beginning 1 hour before surgery. In addition, the eyes in the study group received one drop of topical ofloxacin four times daily for 3 days before surgery. For all patients in both groups, the brow, both upper and lower eyelids, eyelashes, and adjacent forehead, nose, cheek, and temporal orbital area were scrubbed with 5% povidone iodine for 1 minute. In addition, two drops of 5% povidone iodine were placed in the inferior fornix. No intracameral antibiotic was used. In three patients, cataract surgery was performed on both eyes, 14 to 125 days apart. In all three patients, one eye was randomized to the control group, and the other eye to the study group.

Conjunctival cultures were obtained from the operated eye using a cotton swab and immediately were inoculated onto plates containing sheep blood, chocolate, and thioglycolate broth for bacteria and Sabouraud’s agar for fungus. The blood culture media plates were incubated with 5% carbon dioxide to encourage aerobic and microaerophilic bacterial growth. The chocolate agar plates were incubated in an anaerobic bag for isolation of anaerobic bacteria. All cultures were incubated at 37°C for at least 10 days. Culture samples were obtained at time points described in Table 1. At t0, t1, and t2, cultures were obtained without topical anesthetic to optimize bacterial growth by eliminating preservative associated with inhibition of bacterial growth. The individuals who obtained the conjunctival cultures at t0, t1, and t2 were not masked, whereas the surgeons who obtained the culture samples at t3 and t4 were masked as to whether the patient was in the control or study

Figure 2. Maximum colony-forming (CFU) units isolated from blood agar media.

2037

Ophthalmology Volume 109, Number 11, November 2002 Table 1. Description of the Time Points When Cultures Were Taken Time Points

Description

t0

Five days before surgery, before any topical antibiotic Before the application of topical ofloxacin on the day of surgery After 1 hour preoperative antibiotic application and before povidone–iodine scrub After povidone–iodine scrub and immediately before surgery At the conclusion of the surgical procedure

t1 t2 t3 t4

group. The microbiologist responsible for isolating and identifying the bacteria was not masked. All patients underwent anterior chamber aqueous fluid aspiration at the beginning and conclusion of surgery. The results of these cultures will be presented in a separate paper. Statistical analysis was performed using the chi-square test or Mann–Whitney U test with the Analyze-It software program (Analyze-It Software, Leeds, England).

For the blood and chocolate agar culture media, the analysis was made using the Mann–Whitney U test. This nonparametric test was used because the results of the colony counts were not normally distributed. For the blood agar culture media, there was no difference between the control and study groups at t0 (P ⫽ 0.6842). However, at t1, where eyes in the study group had been treated with antibiotic for 3 days, compared with no antibiotic in the control group, significantly fewer bacteria were isolated from the study group compared with the control group (P ⫽ 0.0003). Similarly, a statistically significant difference was found between the two groups at t2 (before povidone iodine scrub) and at t4 (immediately after surgery). No statistically significant difference was found between the two groups at t3 (immediately before surgery). All chocolate agar plates were incubated anaerobically, and there was no statistical difference between the control and study groups at any time point. In most cases in both the control and study groups, there was no bacterial growth in either the blood or chocolate agar media. Therefore, the median colony-forming units (CFU) count for both the control and study groups was zero at all time points except for blood agar culture at t0, where the median was one CFU in both groups. The most commonly isolated bacteria were coagulate-negative Staphylococcus and Propionibacterium acnes. No fungus was isolated from the Sabouraud’s agar plates.

Results

Discussion

The culture results for thioglycolate broth, blood, and chocolate agar are summarized in Table 2. The thioglycolate culture was considered positive if the broth was cloudy within 6 days of incubation. At t0, 28 of 47 samples in the control group (60%) had positive cultures compared with 28 of 41 eyes (68%) in the study group (P ⫽ 0.5313). At t1, 27 of 48 eyes (56%) in the control group had positive cultures, compared with 20 of 44 eyes (45%) in the study group (P ⫽ 0.4088). The change in the percentage of positive cultures from t0 to t1 was 60% to 56% (P ⫽ 0.9042) for the control group and 68% to 56% (P ⫽ 0.0570) for the study group. At time point t3, 20 of 48 eyes (42%) in the control group had positive cultures, compared with only 8 of 43 eyes (19%) in the study group (P ⫽ 0.0314). At time point t4, 16 of 47 eyes (34%) and 6 of 43 eyes (14%) had positive cultures for the control and study groups, respectively (P ⫽ 0.0489). Cultures were not obtained at t0 and t4 from one eye in the control group. In the study group, cultures were not obtained from three eyes at t0 and from one eye at t3 and t4. All P values were calculated using the chi-square test with two-tailed assumption. Figure 1 summarizes the results.

The most common sources of bacteria causing postoperative endophthalmitis are the eyelid and conjunctival areas. Speaker et al2 demonstrated that in 14 of 17 cases of endophthalmitis, the bacterial isolates from the vitreous genetically matched bacterial species isolated from the eyelid, conjunctiva, or nose. Because the incidence of postoperative endophthalmitis is very low, a study would require thousands of patients to show differences in the effectiveness of one treatment group over another in reducing the risk of endophthalmitis. Lowering the number of bacteria on the eyelids and conjunctiva at the time of the surgery, the risk of developing endophthalmitis, may decrease. In this study, we use the frequency of bacteria recovered from the conjunctiva as a surrogate indicator of the risk of endophthalmitis. We assume that the presence of bacteria in the conjunctiva is associated with an increased risk of endophthalmitis. This assumption, however, has not yet been proven. Table 2. Culture Results of Thioglycolate

Thioglycolate Broth Control

Blood Agar

Study

No. of Samples

Maximum No. of Colony-forming Units

Median Count of Colony-forming Units

Time Points

Ratio*

% Eyes with Positive Culture

Ratio*

% Eyes with Positive Culture

P Value

Control Group

Study Group

Control Group

Study Group

Control Group

Study Group

t0 t1 t2 t3 t4

28/47 27/48 24/48 20/48 16/47

60% 56% 50% 42% 34%

28/41 20/44 11/44 8/43 6/43

68% 45% 25% 19% 14%

0.53 0.41 0.02 0.03 0.05

47 48 48 48 48

41 44 44 43 43

130 102 101 3 16

101 4 1 1 0

1 0 0 0 0

1 0 0 0 0

* Numerator represents number of eyes with positive culture and denominator represents total number of eyes cultured.

2038

Ta et al 䡠 Prophylactic Antibiotic for Cataract Surgery Apt et al4 have shown the effectiveness of povidone iodine in reducing the number of bacteria present on the conjunctiva. Several other published studies demonstrated that a combination of povidone iodine and antibiotics significantly decreases conjunctival bacterial flora.5,6 However, there has been a scarcity of data on the use of these antibiotics. This study was designed to determine whether topical ofloxacin was more effective in reducing bacterial flora if used for 3 days before surgery than if used only on the day of surgery. Our results support the use of topical ofloxacin for 3 days before surgery. The most important time points were t3 and t4, immediately before and after surgery, respectively. At t2, t3, and t4, there were significantly fewer positive cultures from thioglycolate broth in the study group compared with the control group (P ⬍ 0.05). The results of blood agar culture media showed that bacterial counts were significantly lower for eyes in the study group compared with those in the control group for time points t1, t2, and t4 (P ⱕ 0.05). Given the nonparametric distribution of the number of colonies isolated, the maximum number of colonies isolated may have more clinical relevance than the mean bacterial counts. Topical ofloxacin alone for 3 days reduced the maximum bacterial counts from 101 CFU at t0 to 4 CFU at t1 and 1 CFU at t2. In the control group, the maximum bacterial counts remained greater than 100 CFU at t1 and t2 (Fig 2). Topical ofloxacin alone for 3 days may be as effective in reducing the bacterial count as povidone iodine. The percentage of positive cultures in thioglycolate broth for the study group at t2, the mean bacterial count on blood agar plates, and the maximum CFU were all lower than the corresponding measurements for the control group at t3, after the use of povidone iodine (Table 3). This suggests that patients who are allergic to iodine may benefit from the use of topical antibiotics 3 days before surgery. We demonstrated a statistically significant difference in the blood agar culture results between study and control groups for t1 and t2, but not for t3. This may be because of the overwhelming effect of 5% povidone iodine in eliminating bacteria, decreasing the number of bacteria isolated in the control group to the same level as that of the study group. However, at the conclusion of surgery, there appeared to be an increase in bacterial counts for the control

Table 3. Effectiveness of Topical Ofloxacin Compared with 5% Povidone–Iodine Control Group at Time Point t3

Study Group at Time Point t2

42% 0.1 CFU 3 CFU

25% 0 CFU 1 CFU

% Positive in thioglycolate broth Mean CFU on blood agar Maximum CFU on blood agar

CFU ⫽ colony-forming units; t3 ⫽ after povidone–iodine scrub; t2 ⫽ before povidone–iodine scrub.

group but not for the study group, resulting in a statistical difference in the number of bacteria isolated between the two groups. This may be because of the usual irrigation and manipulation of the eye during surgery, allowing bacteria that are present at the bases of the eyelashes and conjunctival folds to contaminate the surgical field. The overall number of bacteria present at the bases of eyelashes and conjunctival folds may be lower after 3 days of topical ofloxacin in the study group compared with the control group, resulting in fewer bacteria isolated at the conclusion of surgery (t4) for the study group. Fluoroquinolones comprise a class of broad-spectrum bactericidal antibiotics. The mechanism of action is based on the inhibition of bacterial DNA gyrase. Unlike antiseptics such as povidone iodine, antibiotics do not kill bacteria within seconds on contact, but rather require a longer period of time.11,12 It is conceivable that giving a topical antibiotic for 1 hour before surgery does not allow adequate exposure time to reduce the bacterial count, and therefore, 3 days of antibiotic is more effective. One published study suggests that ciprofloxacin significantly reduces bacterial flora of the conjunctiva within 15 minutes of application.10 However, studies by Pandya et al [Invest Ophthalmol Vis Sci 42(Suppl):254, 2001] and others have shown no difference in the time-kill curve between ofloxacin, ciprofloxacin, and levofloxacin.12,13 Ofloxacin is one of three major topical fluoroquinolones commonly used. Although our results may also apply to ciprofloxacin and levofloxacin, this will require further testing. Our results were consistent with the published data by Isenberg et al,5 who demonstrated that the topical applica-

.Broth, Blood Agar, and Chocolate Agar Chocolate Agar Mean Control Group 11.8 5.0 3.5 0.1 0.5

No. of Samples

Maximum No. of Colony-forming Units

Median Count of Colony-forming Units

Mean

Study Group

P Value

Control Group

Study Group

Control Group

Study Group

Control Group

Study Group

Control Group

Study Group

P Value

5.7 0.1 0 0 0

0.68 ⬍0.001 0.01 0.73 0.05

46 48 48 48 48

41 44 44 43 43

1000 40 6 1 13

16 11 2 2 1

0 0 0 0 0

0 0 0 0 0

24.1 1.6 0.4 0 0.3

1.7 0.5 0.1 0.1 0

0.17 0.45 0.78 0.25 0.51

2039

Ophthalmology Volume 109, Number 11, November 2002 tion of polymixin B sulfate, neomycin sulfate, and gramicidin three times daily for 3 days reduced bacterial counts by 31%.5 Data published by Grimes et al14 also suggested that topical gentamicin was more effective when given for 3 days before surgery than one-half day before surgery.14 One limitation of our study was that the individuals who obtained cultures at t0, t1, and t2 and the microbiologist who examined the specimens were not masked with regard to which antibiotic regimen was used. However, the surgeons who obtained conjunctival specimens at t3 and t4 were masked. In conclusion, our results indicate that the application of topical ofloxacin for 3 days before surgery is more effective than an application given only 1 hour before surgery in reducing the number of positive bacterial cultures at several time points in the perioperative period.

References 1. Aaberg TM Jr, Flynn HW Jr, Schiffman J, Newton J. Nosocomial acute-onset postoperative endophthalmitis survey. A 10-year review of incidence and outcomes. Ophthalmology 1998;105:1004 –10. 2. Speaker MG, Milch FA, Shah MK, et al. Role of external bacterial flora in the pathogenesis of acute postoperative endophthalmitis. Ophthalmology 1991;98:639 – 49 discussion 650. 3. Hale ML. Povidone-iodine in ophthalmic surgery. Ophthalmic Surg 1970;1:9 –13. 4. Apt L, Isenberg S, Yoshimori R, Paez JH. Chemical preparation of the eye in ophthalmic surgery. III. Effect of povidoneiodine on the conjunctiva. Arch Ophthalmol 1984;102:728 –9.

5. Isenberg SJ, Apt L, Yoshimori R, Khwarg S. Chemical preparation of the eye in ophthalmic surgery. IV. Comparison of povidone-iodine on the conjunctiva with a prophylactic antibiotic. Arch Ophthalmol 1985;103:1340 –2. 6. Caldwell DR, Kastl PR, Cook J, Simon J. Povidone-iodine: its efficacy as a preoperative conjunctival and periocular preparation. Ann Ophthalmol 1984;16:577– 80. 7. Apt L, Isenberg SJ, Yoshimori R, Spierer A. Outpatient topical use of povidone-iodine in preparing the eye for surgery. Ophthalmology 1989;96:289 –92. 8. Speaker MG, Menikoff JA. Prophylaxis of endophthalmitis with topical povidone-iodine. Ophthalmology 1991;98:1769 – 75. 9. Binder C, de Kaspar HM, Klauss V, Kampik A. Pra¨ operative Infektionsprophylaxe mit 1%iger Polyvidon-Jod-Lo¨ sung am Beispiel von konjunktivalen Staphylokokken. Ophthalmologe 1999;96:663–7. 10. Snyder-Perlmutter L, Katz HR, Melia M. Effect of topical ciprofloxacin 0.3% and ofloxacin 0.3% on the reduction of bacterial flora on the human conjunctiva. J Cataract Refract Surg 2000;26:1620 –5. 11. Rodeheaver G, Bellamy W, Kody M, et al. Bactericidal activity and toxicity of iodine-containing solutions in wounds. Arch Surg 1982;117:181– 6. 12. Callegan MC, Booth MC, Gilmore MS. In vitro pharmacodynamics of ofloxacin and ciprofloxacin against common ocular pathogens. Cornea 2000;19:539 – 45. 13. Kowalski RP, Pandya AN, Karenchak LM, et al. An in vitro resistant study of levofloxacin, ciprofloxacin, and ofloxacin using keratitis isolates of Staphylococcus aureus and Pseudomonas aeruginosa. Ophthalmology 2001;108:1826 –9. 14. Grimes SR, Mein CE, Trevino S. Preoperative antibiotic and povidone-iodine preparation of the eye. Ann Ophthalmol 1991;23:263– 6.

Discussion by Eric D. Donnenfeld, MD The basic principles of antibiotic prophylaxis are: (1) to deliver an active agent effective against infecting micro-organisms; (2) to have adequate contact time between the drug and the infecting microbe, because most antibiotics require bacterial replication to perform complete killing of the organism; (3) to avoid or minimize toxic side effects; (4) to minimize the risk of development of resistant organisms; and (5) to control cost. Ta et al have investigated quantitatively and qualitatively and have compared the effect of preoperative topical ofloxacin prophylaxis begun 3 days before surgery with prophylactic therapy begun 1 hour before surgery on conjunctival cultures. Antibiotic prophylaxis for prevention of endophthalmitis in cataract surgery is one of the most controversial areas in ophthalmology because of the ocular morbidity of endophthalmitis coupled with the paucity of clinical studies that conclusively show an effective technique to reduce its occurrence. In fact, the most recent study documenting the advantages of the use of prophylactic antibiotics was published in 1974 by Allen and Mangiaracine.1 They showed a 90% reduction in the risk of endophthalmitis when

From the Ophthalmic Consultants of Long Island Centre, New York. Address correspondence to Eric D. Donnenfeld, MD, Ryan Medical Arts Building, Suite 402, 2000 North Village Avenue, Rockville Centre, NY 11570.

2040

prophylactic antibiotics were used. The only other prospective study that has documented efficacy in reducing endophthalmitis is the study by Speaker and Menikoff,2 which showed that preoperative topical 5% povidone iodine significantly reduced the risk of endophthalmitis. With these limitations notwithstanding, and with the inability to perform a prospective controlled clinical series in the medical–legal environment of today, we rely on indirect clinical studies to suggest that antibiotics reduce the risk of endophthalmitis. In the study presented, Ta et al have shown that 3 days of prophylactic, commercially available ofloxacin significantly reduces the qualitative incidence of positive bacterial cultures before and after povidone iodine application and on completion of cataract surgery. They have shown a marked decrease in quantitative colony count on blood agar before the addition of povidone iodine and, perhaps most importantly, that the maximum number of colony-forming units for the experimental group was dramatically lower than that of the control group at all time intervals. Previous studies3,4 have shown a high incidence of positive anterior chamber bacterial cultures on the conclusion of cataract surgery. Clearly, the body’s immune system is able to sterilize a small bacterial inoculum. In fact, rabbit models have shown that an inoculation of 30,000 Staphylococcus epidermidis bacteria are necessary to induce endophthalmitis and that smaller inoculums autosterilize.5 This suggests that endophthalmitis will occur in