Table 1.
Topical antibiotics in adults and children with suspected or confirmed bacterial conjunctivitis: results of RCTs.
| Intervention | Number of participants | Age of participants | Proportion culture-positive | Microbiological cure rate | Clinical cure rate | Adverse effects |
| Topical antibiotics versus placebo | ||||||
| Suspected bacterial conjunctivitis | ||||||
| Norfloxacin 0.3% every 2 hours while awake, for 1 day, then 4 times daily for 5 to 6 more days | 284 adults | 18 years and over | With or without positive cultures | At 2 to 3 days: 65% with norfloxacin v 26% with placebo; P <0.01At 5 to 7 days (excluding coagulase-negative Staphylococcus): 74% with norfloxacin v 42% with placebo; P <0.01 | At 5 days: 88% with norfloxacin v 72% with placebo; P <0.01 | Minor events (including chemosis and burning): 4% with norfloxacin v 7% with placebo; significance not reported |
| Chloramphenicol 0.5% 1 drop every 2 hours while awake, for 1 day, then 4 times daily until 48 hours after infection has resolved | 326 children | 6 months to 12 years | 78% | Microbiological cure at 7 days: 40% with chloramphenicol v 23% with placebo; ARI 17%, 95% CI 5.5% to 28.1% Microbiological cure or improvement at 7 days: 65% with chloramphenicol v 55% with placebo; ARI +9.6%, 95% CI –2.5% to +21.7% | At 3 days: 39% with chloramphenicol v 33% with placebo; ARI +6.2%, 95% CI –4.3% to +16.5%At 7 days: 86% with chloramphenicol v 79% with placebo; ARI +7.4%, 95% CI –0.9% to +15.6% | 2% with chloramphenicol v 2% with placebo; ARI 0%, 95% CI –2.9% to +2.9% |
| Fusidic acid gel 10 mg/g 1 drop 4 times daily until 1 day after signs and symptoms disappear | 181 adults | 18 years and over | 34% | At 7 days: 76% with fusidic acid v 41% with placebo; ARI 34.8%, 95% CI 9.3% to 60.4% | At 7 days: 62% with fusidic acid v 59% with placebo; ARI +2.8%, 95% CI –13.5% to +18.6% | 14% with fusidic acid v 3% with placebo; ARI 10.4%, 95% CI 1.6% to 19.1% |
| Chloramphenicol eye drops (immediate, every 2 hours for 2 days, then 4 times daily), chloramphenicol eye drops (delayed, same regimen, prescriptions available from surgery up to 3 days after consultation at patient's/parent's discretion), no antibiotics | 307 adults and children | Over 1 year, mean age 27 years | 50% | Not reported | Mean symptom score (days 1–3 after consultation): 1.9 with immediate antibiotics v 2.1 with no antibiotics; P = 0.2 Mean symptom score (days 1–3 after consultation): 2.0 with delayed antibiotics v 2.1 with no antibiotics; P = 0.4 | 1 person in immediate-antibiotic group was admitted 11 days post-consultation for orbital cellulitis |
| Besifloxacin 0.6% twice a day for 3 days | 202 adults and children | 1 year and over | 54% | At days 4/5: 87% with besifloxacin v 57% with vehicle; P <0.001 At days 6/7/8: 87% with besifloxacin v 70% with vehicle; P = 0.038 | At days 4/5: 70% with besifloxacin v 38% with vehicle; P <0.001 At days 6/7/8: 74% with besifloxacin v 66% with vehicle; P = NS Individual clinical outcomes of ocular discharge (83% with besifloxacin v 54% with vehicle) and conjunctival injection (77% with besifloxacin v 47% with vehicle) also favoured besifloxacin at visit 2, but not at visit 3 | No difference in adverse events in all treated eyes: 6% with besifloxacin v 11% with vehicle; all were mild to moderate severity (non-specific conjunctivitis, bacterial conjunctivitis, and installation-site pain); non-ocular adverse events similar between groups and not considered related to treatments |
| Confirmed bacterial conjunctivitis | ||||||
| Levofloxacin 0.5% | 249 recruited; 117 people included in efficacy analysis | 2 to 91 years | 117 culture-positive and included in per-protocol cohort | At "end point", defined as the last evaluable observation (up to day 10 post-treatment): 90% with levofloxacin v 53% with placebo; P <0.001 | At "end point", defined as the last evaluable observation (up to day 10 post-treatment): 77% with levofloxacin v 60% with placebo; P = 0.026 | No significant difference between levofloxacin and placebo in transient burning (2.4% of 124 people); transiently decreased vision (2.4% of 124 people) with levofloxacin |
| Moxifloxacin | 73 recruited; number included in efficacy analysis not reported; unclear whether analysis was restricted only to people with culture-positive bacterial conjunctivitis | 1 to 89 years | 51 culture-positive | After about 1 week of treatment: 78% with moxifloxacin v 39% with placebo; P = 0.005 | After about 1 week of treatment: 93% with moxifloxacin v 63% with placebo; P = 0.009 | All adverse events reported as not serious |
| Polymyxin B sulphate 10,000 U/g plus bacitracin 500 U/g (in ointment) 4 times daily for 7 days | 66 | 1 month to 18 years | Cultures positive for Haemophilus influenzae or Streptococcus pneumoniae | Eradicated at 3 to 5 days: 71% with polymyxin B sulphate–bacitracin v 19% with placebo; P <0.001Eradicated at 8 to 10 days: 79% with polymyxin B sulphate–bacitracin v 31% with placebo; P <0.001 | Cured at 3 to 5 days: 62% with polymyxin B sulphate–bacitracin v 28% with placebo; P <0.02At 8 to 10 days: 91% with polymyxin B sulphate–bacitracin v 72% with placebo; P >0.05; NS | Allergic reaction to topical polymyxin B sulphate–bacitracin in initial group of participants |
| Polymyxin B sulphate plus bacitracin in people taking systemic antibiotics: amoxicillin, trimethoprim–sulfamethoxazole, cefaclor, or penicillin (subgroup analysis of RCT described above) | 18 | 1 month to 18 years | Cultures positive for H influenzae or S pneumoniae | 72% with polymyxin B sulphate–bacitracin eradicated at 3 to 5 days, and 78% at 8 to 10 days. In people receiving systemic antibiotics, no significant difference in microbiological cure between adding polymyxin B sulphate–bacitracin and adding placebo (reported as NS; P value not reported) | 83% cured at 3 to 5 days; 100% cured at 8 to 10 days. In people receiving systemic antibiotics, there was no significant difference in clinical cure between adding polymyxin B sulphate–bacitracin and adding placebo (reported as NS; P value not reported) | Allergic reaction to topical polymyxin B sulphate–bacitracin in initial group of participants |
| Ciprofloxacin 0.3% every 2 hours while awake on days 0 to 1, then every 4 hours while awake for 1 to 2 more days | 177 | Age not specified | Culture-positive | Eradicated or reduced at 3 days: 132/140 (94%) with ciprofloxacin v 22/37 (59%) with placebo; RR 1.59; P <0.001 | Not reported | Adverse effects not assessed in the RCT |
| Ofloxacin 0.3% 6 times daily for 2 days | 132 | Age not specified | Culture-positive | At 2 days: 72% with ofloxacin v 35% with placebo; P <0.001 | Improved at 2 days: 64% with ofloxacin v 22% with placebo; P <0.001 | Adverse effects not assessed in the RCT |
| Besifloxacin 0.6% v vehicle only (placebo) 1 drop 3 times daily for 5 days | 118 (269 in "safety" population who got drug for clinical diagnosis) | 1 to 92 years | 44% | At day 4: 90% with besifloxacin v 47% with vehicle; P <0.001At day 8: 88% with besifloxacin v 60% with vehicle; P <0.001 | At day 4: 33% with besifloxacin v 17% with vehicle; P = NSAt day 8: 73% with besifloxacin v 43% with vehicle; P <0.001 | Mild–moderate severity adverse events common (50% with besifloxacin v 53% with vehicle); 1 preseptal cellulitis in vehicle group thought unrelated to study |
| Besifloxacin 0.6% v vehicle only (placebo) 1 drop 3 times daily for 5 days | 390 culture-positive for efficacy analysis (957 enrolled for clinical diagnosis analysed for safety) | 10 months to 98 years | 41% | At day 5: 92% with besifloxacin v 59% with vehicle; P <0.0001At days 8 to 9: 88% with besifloxacin v 72% with vehicle; P <0.0001 | At day 5: 45% with besifloxacin v 33% with vehicle; P = 0.0084At days 8 to 9: 84% with besifloxacin v 69% with vehicle; P = 0.0011 | More conjunctivitis (non-specific and bacterial) in vehicle-only group (14% with besifloxacin v 9% with placebo; P = 0.0047); more pruritus in besifloxacin group (1% with besifloxacin v 0.3% with placebo; P = 0.03) and viral conjunctivitis (0.7% with besifloxacin v 0% with placebo; P = 0.02) |
| Azithromycin 1% 1 drop twice daily on days 1 and 2, then once daily on days 3 to 5 v vehicle-only placebo dosed in same manner | 685 enrolled, 630 completed, 279 analysed in per-protocol analysis | 1 to 96 years | 41% | At visit 3 (day 6 or 7): 89% with azithromycin v 66% with vehicle; P <0.001; difference 22.0%, 95% CI 12.7% to 31.4% | At visit 3: 63% with azithromycin v 50% with vehicle; P <0.03; difference 13.4%, 95% CI 1.9% to 25.0% | Similar between groups in frequency and magnitude |
| Topical antibiotics versus each other | ||||||
| Suspected bacterial conjunctivitis | ||||||
| Chloramphenicol 0.5% drops v tobramycin 0.3% | 50 | 8 to 81 years | 36% culture-positive for bacteria (2% positive for Candida) | Not reported | No significant difference between chloramphenicol and tobramycin in clinical scores assessed by patients or investigators (P >0.05) | 0% with chloramphenicol v 8% with tobramycin had irritation; statistical analysis not reported |
| Ciprofloxacin 0.3% drops v tobramycin 0.3% drops; regimen unspecified (abstract reviewed, but full paper unavailable) | 40 | Age unspecified | Not all culture-confirmed | Eradication: 80% with ciprofloxacin v 95% with tobramycin; reported as NS | Cure: 95% with ciprofloxacin v 95% with tobramycin; reported as NS | Adverse events (burning, bitter taste, pruritus, punctate epithelial erosions: 20% with ciprofloxacin v 35% with tobramycin |
| Fusidic acid 1% viscous drops twice daily v rifamycin 1% drops 4 times daily | 163 | Adults and children | 72% to 75% culture-positive | Not reported in each group separately | Cure: 87% with fusidic acid v 89% with rifamycin; P = 0.71; median: 7 days with fusidic acid v 6 days with rifamycin; P = 0.31 | Burning and bad taste: 4/74 (5%) with fusidic acid v 13/77 (17%) with rifamycin; reports of allergy with rifamycin |
| Fusidic acid 1% viscous drops twice daily v norfloxacin 0.3% drops 4 times daily | 400 | Over 1 year | 34% culture-positive | Not reported | Success of treatment as assessed by investigator after 7 days' treatment: 91% with fusidic acid v 93% with norfloxacin; P = 0.49 | Bad taste: 6% with fusidic acid v 20% with norfloxacin; P = 0.001; stinging: 37% with fusidic acid v 50% with norfloxacin; P = 0.007 |
| Fusidic acid 1% viscous drops twice daily v chloramphenicol 0.5% drops 4-hourly | 541 | Over 1 year | 17% culture-positive | Not reported | Success of treatment, assessed by investigator: 96% with fusidic acid v 97% with chloramphenicol cured; P = 0.56Complete absence of symptoms: 71% with fusidic acid v 77% with chloramphenicol; P = 0.14 | Bad taste: 11% with fusidic acid v 37% with chloramphenicol; P = 0.001 |
| Fusidic acid 1% viscous drops twice daily after loading dose v chloramphenicol 0.5% drops 6 times daily after loading dose | 340 | Adults and children (ratio not specified) | 161/340 (47%) culture-positive | Not reported | >90% cured/improved; median: 6.6 days with fusidic acid v 6.2 days with chloramphenicol; no significant difference between fusidic acid and chloramphenicol | Itching, burning, blurred vision, bad taste: 31% with fusidic acid v 16% with chloramphenicol |
| Fusidic acid 1% suspension in carbomer gel twice daily after loading dose v chloramphenicol 0.5% drops 5 to 6 times daily after loading dose | 250 | 221 adults (16–89 years), 29 children (1–14 years) | Not all culture-confirmed | Not reported | Cured: 84% with fusidic acid v 81% with chloramphenicol (mean: 3.3 days with fusidic acid v 3.6 days with chloramphenicol); P = NS | Mild to moderate itching, stinging, local discomfort: 5% with fusidic acid v 14% with chloramphenicol |
| Fusidic acid viscous drops 1% twice daily for 5 to 7 days v chloramphenicol 1% ointment 3-hourly | 505 recruited; 16 lost to follow-up | 1 to 90 years | 27% of 486 culture-positive for pathogenic bacteria | Not reported | 83% with fusidic acid v 84% with chloramphenicol; P = NS | Smarting, irritation, stinging, red eye, blurred vision: 15% with fusidic acid v 11% with chloramphenicol; treatment discontinuation because of adverse effects greater with chloramphenicol (P <0.01) |
| Lomefloxacin 0.3% drops twice daily v norfloxacin 0.3% 4 times daily | 145 | Age not specified | 27% culture-positive | No significant difference in reduction of bacterial counts between lomefloxacin and norfloxacinBy day 7 to 9, colony count score reduced by 96% with lomefloxacin v 85% with norfloxacin (P = 0.47) | No significant difference in reduction of signs and symptoms at 7 to 9 days between lomefloxacin and norfloxacinClinical scores reduced by 96% with lomefloxacin v 90% with norfloxacin (P >0.4) | 12 with lomefloxacin v 14 with norfloxacin (more burning with norfloxacin) |
| Lomefloxacin 0.3% drops twice daily after loading dose v chloramphenicol 0.5% drops 5 times daily after loading dose | 191 | 16 to 85 years | 96/191 culture-positive | No significant difference between lomefloxacin and chloramphenicol by 3 to 5 days, 0 colonies in 79% with lomefloxacin v 80% with chloramphenicol; no significant difference in colony count scores by 3 to 5 days (P = 0.97) or at days 7 to 9 (P = 0.12) | No significant difference between lomefloxacin and chloramphenicol in the cumulative score of signs and symptoms in people with bacteriologically confirmed (at 3–5 days, P = 0.83; at 7–9 days, P = 0.18) or clinically diagnosed (3–5 days, P = 0.54; 7–9 days, P = 0.63) bacterial conjunctivitis | Good to excellent tolerance rating |
| Lomefloxacin 0.3% drops twice daily v gentamicin 0.3% drops 4 times daily after loading dose | 66 | 8 to 80 years | 46% culture-positive | Most positive cultures were eradicated by days 3 to 5, with no significant difference between lomefloxacin and gentamicin. By days 3 to 5, positive cultures eradicated in 21/32 with lomefloxacin v 27/32 with gentamicin (P = 0.91) | In people with culture-positive bacterial conjunctivitis, no significant difference in clinical scores at 7 to 9 days between lomefloxacin and gentamicin (reduced by 82% with lomefloxacin v 78% with gentamicin; P = 0.58) In people with clinically diagnosed bacterial conjunctivitis, clinical scores reduced by 78% with lomefloxacin v 73% with gentamicin (P = 0.58) at 7 to 9 days | Adverse events: 1 with lomefloxacin v 3 with gentamicin (more burning with gentamicin) |
| Lomefloxacin 0.3% twice daily v tobramycin 0.3% 4 times daily | 99 recruited, 92 completed | Mean age 42 years; range 11 to 80 years | About 50% | At days 1 and 2: 48% with lomefloxacin v 55% with tobramycinAt days 7 and 8: 23% with lomefloxacin v 36% with tobramycin (reported as NS; P value not reported) | Not reported | Similar rates and duration of burning sensation after instillation in both groups |
| Lomefloxacin 0.3% drops v fusidic acid 1% gel twice daily after loading dose | 45 | Adults and children (ratio not specified) | 81% culture-positive | Eradicated at days 3 to 5: 8/15 (53%) with lomefloxacin v 4/16 (25%) with fusidic acid; P = 0.075 | No significant difference between lomefloxacin and fusidic acid in reduction of signs and symptoms (reported as NS, absolute results presented graphically) | Significantly more people using fusidic acid had burning (11% with lomefloxacin v 48% with fusidic acid; P = 0.009) |
| Tobramycin 0.3% drops, 1 to 2 drops 4 to 6 times daily v fusidic acid 1% viscous drops, 1 drop twice daily for 7 days | 494 recruited; 487 treated; 8 people lost to follow-up; information provided only for subgroup with pathogenic bacteria | 2 to 85 years; cohort was subdivided into 2 groups (2–9 years and over 9 years) | 66% culture-positive, but 70% of culture-positive people had normal flora on quantitative microbiology | No significant difference between fusidic acid and tobramycin after 7 days' treatment (81% with fusidic acid v 88% with tobramycin; P = 0.34) | No significant difference in signs and symptoms at 7 days In children aged 2 to 9 years: 77% with fusidic acid v 83% with tobramycinIn people aged over 9 years: 76% with fusidic acid v 73% with tobramycin (reported as NS; P value not reported) | Fusidic acid 4% (tearing, burning, irritation, stinging, allergic reaction, conjunctival injection), tobramycin 2% (irritation, pain, red eye, photosensitivity, discharge; P value not reported); 2 people withdrawn from each treatment group because of adverse effects |
| Combination of trimethoprim (5 mg/g) and polymyxin B sulphate (10,000 U/g) v chloramphenicol (10 mg/g) as ointment 4 times daily | 42 | Adults and children (ratio not specified) | 55% culture-positive | Eradicated: 13/16 (81%) with trimethoprim–polymyxin B sulphate v 4/9 (44%) with chloramphenicol; P value not reported | >90% reduction in signs and symptoms at day 10: 88% with trimethoprim–polymyxin B sulphate v 71% with chloramphenicol >50% reduction: 100% with trimethoprim–polymyxin B sulphate v 94% with chloramphenicol (P = NS) | 3 people using trimethoprim–polymyxin B sulphate reported stinging, grittiness, conjunctival hyperaemia, or lid oedema |
| Combination of trimethoprim (1 mg/mL) and polymyxin B sulphate (10,000 U/mL) v chloramphenicol drops 6 times daily for 7 days | 40 | 8 to 70 years (ratio not specified) | 95% culture-positive | Not reported | No significant difference between trimethoprim–polymyxin B sulphate and chloramphenicol in reduction in signs/symptoms score at 7 days (56% with trimethoprim–polymyxin B sulphate v 57% with chloramphenicol; reported as NS; P value not reported) | Adverse effects not assessed in the RCT |
| Combination of trimethoprim (1 mg/mL) plus polymyxin B sulphate (10,000 U/mL) drops v chloramphenicol (5 mg/mL) 4 times daily. Multicentre trial with 2 separate comparisons (other comparison reported below) | 130 | Adults and children (ratio not specified) | 43% culture-positive | Eradicated: 19/24 (79%) with trimethoprim–polymyxin B sulphate v 21/26 (81%) with chloramphenicol | >90% reduction in signs and symptoms at days 10 to 14: 74% with trimethoprim–polymyxin B sulphate v 54% with chloramphenicol (P = NS)>50% reduction: 95% with trimethoprim–polymyxin B sulphate v 85% with chloramphenicol (P = NS) | 4 withdrawals from study because of stinging v 3 withdrawals because of allergic reaction |
| Combination of trimethoprim (1 mg/mL) plus polymyxin B sulphate (10,000 U/mL) drops v combination of polymyxin B sulphate (5000 U/mL) plus neomycin (1700 U/mL) plus gramicidin (25 U/mL) 4 times daily. Multicentre trial with 2 separate comparisons (other comparison reported above) | 100 | Adults and children (ratio not specified) | 43% culture-positive | Eradicated: 15/27 (56%) with trimethoprim–polymyxin B sulphate v 18/33 (55%) with polymyxin B sulphate | >90% reduction in signs and symptoms at days 10 to 14: 80% with trimethoprim–polymyxin B sulphate v 68% with polymyxin B sulphate–neomycin–gramicidin (P >0.05); >50% reduction in signs and symptoms: 96% with trimethoprim–polymyxin B sulphate v 88% with polymyxin B sulphate–neomycin–gramicidin (P = NS) | See trimethoprim–polymyxin B sulphate group adverse events above; 1 withdrawal from polymyxin B sulphate group because of periorbital oedema |
| Trimethoprim–polymyxin B sulphate drops v neomycin–polymyxin B sulphate–gramicidin drops 6 times daily | 48 | Adults and children (ratio not specified) | 46% culture-positive | Eradicated: 8/8 (100%) with trimethoprim–polymyxin B sulphate v 12/14 (86%) with neomycin–polymyxin B sulphate–gramicidin | No significant difference in symptoms and signs after 10 days' treatment between trimethoprim–polymyxin B sulphate and neomycin–polymyxin B sulphate–gramicidin (reported as NS; P value not reported; absolute results tabulated) | Adverse effects not assessed in the RCT |
| Combination of trimethoprim (5 mg/g) and polymyxin B sulphate (10,000 U/g) v chloramphenicol (10 mg/g) as ointment 3 or 4 times daily (4 separate RCTs of this comparison reported in this article) | 448 | Adults and children (ratio not specified) | 32% to 72% culture-positive | Not reported | Trial 1: 73% with trimethoprim–polymyxin B sulphate v 67% with chloramphenicol cure (P >0.1)Trial 2: 65% with trimethoprim–polymyxin B sulphate v 42% with chloramphenicol cure (P = 0.1)Trial 3: 80% with trimethoprim–polymyxin B sulphate v 64% with chloramphenicol cure (P >0.1)Trial 4: 37% with trimethoprim–polymyxin B sulphate v 50% with chloramphenicol cure (P >0.1) (at 10 days) | 22 with trimethoprim–polymyxin B sulphate v 12 with chloramphenicol people (stinging, swollen lids, irritation, tearing) |
| Moxifloxacin 0.5% 1 drop 3 times daily v trimethoprim 1%–polymyxin B sulphate 10,000 IU 1 drop 4 times daily*Note: trimethoprim–polymyxin B sulphate dose is the lowest recommended dose for the condition for adults and is lower than that used in most of the other studies of trimethoprim–polymyxin B sulphate reviewed here. Manufacturer has no recommended paediatric dose | 56 | 1 month to 18 years | 68/84 (81%) eyes | Microbiological cure rate at 48 hours was broken down by pathogen isolated and showed significant differences favouring moxifloxacin for all bacterial pathogens | Culture-positive eyes at 48 hours: clinical cure rate 81% with moxifloxacin v 44% with trimethoprim–polymyxin B sulphate; P = 0.001All eyes at 48 hours: clinical cure rate 88% with moxifloxacin v 44% with trimethoprim–polymyxin B sulphate; P = 0.001*Note: unit of analysis was not the unit of randomisation | No treatment-related adverse events; 1 episode otitis media in moxifloxacin group and 1 episode respiratory syncytial virus infection in trimethoprim–polymyxin B sulphate group |
| Azithromycin 1.5% 1 drop twice a day for 3 days v tobramycin 0.3% 1 drop every 2 hours for 2 days, then every 4 times a day for 5 days | 150 | 4 to 17 years | 58% | Day 3: 94% with azithromycin v 76% with tobramycin; P <0.01Day 9: 87% with azithromycin v 90% with tobramycin; P <0.01 by bootstrap estimation of means | ITT analysis: overall, day 3: "similar results" to "microbiologically validated" ITT group (below); day 9: 78% with azithromycin v 81% with tobramycin"Microbiologically validated" ITT (those with positive cultures): day 3: 48% with azithromycin v 27% with tobramycin (significant difference by bootstrap estimation of means; P <0.001); day 9: 80% with azithromycin v 82% with tobramycin | 1 patient in azithromycin group had itching, burning, stinging, foreign body sensation, and blurry vision; 2 patients in tobramycin group had itching, burning, stinging, and/or stickiness |
| Besifloxacin 0.6% drops v moxifloxacin 0.5% drops both given 3 times a day for 5 days | 1161 in ITT and safety group, 533 in culture-confirmed group (primary outcome population) | 11 months to 100 years | 533/1161 | Modified ITT (mITT, culture-confirmed) population: day 5: 93% with besifloxacin v 91% with moxifloxacin (non-inferiority, P = 0.1238); day 8: 87% with besifloxacin v 85% with moxifloxacin (non-inferiority, P = 0.0608) ITT population: data not provided; according to authors, data showed similar non-inferiority | mITT (culture-confirmed) population: day 5: 58% with besifloxacin v 59% with moxifloxacin (non-inferiority, P = 0.6520; day 8: 85% with besifloxacin v 84% with moxifloxacin (non-inferiority, P = 0.5014) Investigator's "global assessment of clinical response", mITT population: day 5: 56.7% with besifloxacin v 57.3% with moxifloxacin; day 8: 84.9% with besifloxacin v 84.7% with moxifloxacinITT population: data not provided: according to authors, data showed similar non-inferiority | At least 1 mild–moderate ocular adverse event: 12% with besifloxacin and 14% with moxifloxacin, P = 0.2238; "eye irritation": 0.3% with besifloxacin v 1.4% with moxifloxacin, P = 0.0201; other mild–moderate adverse events: non-significant (conjunctivitis, bacterial conjunctivitis, blurred vision, eye pain), 2 serious adverse events (1 in each group) considered unrelated to treatment |
| Confirmed bacterial conjunctivitis | ||||||
| Ciprofloxacin 0.3% drops 4-hourly while awake after loading dose v tobramycin 0.3% drops 4-hourly while awake after loading dose | 241 | Age unspecified | Culture-confirmed | Eradication or reduction: 94% with ciprofloxacin v 92% with tobramycin (P = 0.5) | Not reported | Adverse effects not assessed in the RCT |
| Ciprofloxacin 0.3% drops 2-hourly for 2 days then 4 times daily for 5 more days v tobramycin drops 2-hourly for 2 days then 4 times daily for 5 more days | 257 (only 141 evaluated for efficacy, but all evaluated for safety) | 0 to 12 years | 100% culture-positive | Eradicated: 90% with ciprofloxacin v 84% with tobramycin; P = 0.29 | Cured by investigator assessment on day 7: 87% with ciprofloxacin v 90% with tobramycin (P = 0.6) | 3 people in each group had adverse effects (dry eye, pruritus, lid oedema, leukoderma, hyperaemia; significance not calculated); 2 people using tobramycin withdrew as a result |
| Fusidic acid 1% gel v chloramphenicol 0.5% drops 4 to 6 times daily for 7 days | 139 (114 with fusidic acid v 25 with chloramphenicol) (248 total, but only the 139 culture-positive patients used to calculated success rates) | Up to 15 years | 100% culture-positive (56% of the total 248) | Not reported (resistance: 16% with fusidic acid v 55% with chloramphenicol; statistical analysis not provided) | 85% with fusidic acid v 48% with chloramphenicol; P <0.0001 | No adverse events associated with treatment reported by participants |
| Levofloxacin 0.5% v ofloxacin 0.3% | 423 recruited; 208 people included in efficacy analysis | 1 to 91 years | 100% | At final visit (6–10 days after start of treatment): 89% with levofloxacin v 80% with ofloxacin; P = 0.034 At end point (defined as last observation, up to and including day 10 after start of treatment): 90% with levofloxacin v 81% with ofloxacin; P = 0.038 | Cured at end point (defined as last observation, up to and including day 10 after start of treatment): 76% in each group; P >0.05 | Burning: 1.45% with levofloxacin v 0.97% with ofloxacin; other adverse effects not examined |
| Levofloxacin 0.3% v ofloxacin 0.3% | 132 (72 with culture-confirmed bacterial conjunctivitis) | 18 to 65 years | 100% | Not reported | Similar cure rates at end of study: 97% with levofloxacin v 94% with ofloxacin either completely or obviously improved; P value not reportedNo significant difference in number of days until improved (mean: 4.89 days with levofloxacin v 5.13 days with ofloxacin; P >0.05) | 2 people using levofloxacin and 1 using ofloxacin had slight irritation |
| Lomefloxacin 0.3%, 1 drop 2-hourly on day 1 then twice daily for 1 week v ofloxacin 0.3% 4 times daily for 1 week | 45 entered, 40 completed | Mean 30 years; range 1 to 78 years | 100% | Not reported | 88% with lomefloxacin v 75% with ofloxacin; P <0.08 | 1 person in each group reported burning sensation after instillation |
| Netilmicin v gentamicin, 1 to 2 drops 4 times daily for up to 10 days | 209 recruited; 121 analysed, all of whom were culture-positive at baseline | Mean (± SD) 49 ± 19 years | 100% of those analysed were culture-positive | Netilmicin significantly more effective than gentamicin at 5 days (P = 0.001) and 10 days (P = 0.037); absolute results presented graphically | Netilmicin significantly more effective than gentamicin at 3 days (P = 0.037), 5 days (P = 0.001), and 10 days (P = 0.001); absolute results presented graphically | 2% with netilmicin v 4% with gentamicin (adverse events included redness, itching, and burning) |
| Trimethoprim hemisulphate 1.0 mg/mL plus polymyxin B sulphate 10,000 U/mL v gentamicin sulphate 3 mg/mL v sulfacetamide 100 mg/mL; all for 10 days | 158 | 2 months to 22 years | 100% culture-positive for H influenzae or S pneumoniae | At 2 to 7 days after treatment: 83% with trimethoprim–polymyxin B sulphate v 68% with gentamicin v 72% with sulfacetamide; P = NS | At 2 to 7 days after treatment: 84% with trimethoprim–polymyxin B sulphate v 88% with gentamicin v 89% with sulfacetamide; P = NS | Similar safety profiles |
| Levofloxacin 0.5% 1 drop 3 times daily v levofloxacin 0.5% 1 drop 2-hourly on days 1 and 2 then 1 drop 4-hourly on days 3 to 5 (usual dosing) | 86 (119 originally enrolled, but 27 had negative bacteriological results) | 18 to 70 years | 72% | 93% with 3-times-daily dosing v 96% in usual dosing; P = 1.00, NS | 85% with 3-times-daily dosing v 92% in usual dosing; P = 0.48, NS | The RCT reported no adverse events in the studied groups |
| Azithromycin 1.5% 1 drop twice daily for 3 days v tobramycin 0.3% 1 drop 2-hourly for 2 days then 4 times daily for 5 days | 1043 patients randomised (ITT set), 1015 in safety set (all "evaluable" participants who got medication), 521 in modified ITT set (culture-positive), and 417 in per-protocol set (no protocol deviations) | 4 days to 87 years | 50% (52% for azithromycin and 48% for tobramycin) | Bacterial resolution on worse eye only (or right eye if equal severity) in per-protocol set:At day 3: 85% with azithromycin v 84% with tobramycin (difference +1.4%, 95% CI –5.3% to +8.3%)At day 9: 93% with azithromycin v 95% with tobramycin (difference –1.8%, 95% CI –6.6% to +3.0%) | Clinical cure at 9 days:Per-protocol set: 88% with azithromycin v 89% with tobramycin (ARD +1.6%, 95% CI –7.5% to +4.4%)Modified ITT set: 86% with azithromycin v 86% with tobramycin (ARD +0.5%, 95% CI –6.6% to +5.8%)ITT set: 85% azithromycin v 85% tobramycin (ARD: +0.5%, 95% CI –3.8% to +4.9%) | Adverse events mild to moderate only: 3/508 (0.5%) reported effects related with azithromycin (burning, foreign body sensation) and 2 discontinued the study; 1/502 (0.1%) reported discharge with tobramycin |
| Topical versus oral antibiotics | ||||||
| Suspected bacterial conjunctivitis | ||||||
| Polymyxin B sulphate–bacitracin 4 times daily for 7 days v oral cefixime 8 mg/kg daily for 3 days | 80 children | 2 months to 6 years | 70% culture-positive | Bacteriological failure at 3 days: 18% with polymyxin B sulphate–bacitracin v 38% with cefixime; P = 0.07 | Not stated but difference reported as NS | Adverse effects not assessed in the RCT |
Antibiotic dosing ranges in this table may vary from the usual clinical recommendations for mild conjunctivitis. However, they are within the accepted ranges for clinician-directed treatment of conjunctivitis based on severity as recommended in major pharmacotherapeutic reference databases.ARD, absolute risk difference; ARI, absolute risk increase; ITT, intention to treat; NS, not significant; SD, standardised difference.