Polymyxin B sulfate and trimethoprim

What it is

This is an eye-drop combining two antibacterial drugs in one sterile solution. Trimethoprim blocks a bacterial enzyme needed to make folate, and polymyxin B damages the outer membrane of Gram-negative bacteria. The label records that after two drops in the eye, serum levels measured in 11 volunteers peaked at roughly 0.03 mcg/mL trimethoprim and 1 unit/mL polymyxin B, so very little of either drug reaches the bloodstream. Systemic polymyxin B for injection is a different product with a different route and a very different safety profile, and nothing in this entry about the eye drop should be read across to it.

What the research says

The evidence base specific to this combination is thin and old. For the broader question the drops are prescribed to answer - does a topical antibiotic help acute bacterial conjunctivitis - the best evidence is a 2023 Cochrane review of 21 placebo-controlled trials, which found a modest benefit of topical antibiotics as a class, and also found that over half of placebo-treated participants got better without any antibiotic. The trials of this particular combination are small, mostly from the 1980s, and compared it with other antibiotics rather than with placebo; the one modern head-to-head trial that measured speed of recovery found it slower than moxifloxacin.

Evidence grade: Limited evidence.

How it works

Drug class: Fixed-combination topical ocular antibacterial: a cyclic lipopeptide (polymyxin B sulfate) plus a dihydrofolate reductase inhibitor (trimethoprim)

Trimethoprim binds and reversibly inhibits bacterial dihydrofolate reductase, blocking the conversion of dihydrofolic acid to tetrahydrofolic acid, which bacteria need to build nucleic acids and proteins. Polymyxin B is a cyclic lipopeptide that increases the permeability of the bacterial cell membrane by interacting with its phospholipids, which kills Gram-negative organisms including Pseudomonas aeruginosa. (Source 1)

What it is used for

• A 2023 Cochrane review of topical antibiotics as a class against placebo found a modest improvement in clinical cure (RR 1.26) at moderate certainty, with 55.5% of placebo-treated participants resolving anyway. Trials of this specific combination are small, mostly pre-1990, and used other antibiotics rather than placebo as the comparator. Evidence: limited. (Source 2)
• The label lists blepharoconjunctivitis as an indication, but we found no randomised trial of this combination against placebo for that condition in the searches we ran. The label is a regulatory position, not an outcome trial. Evidence: unknown. (Source 3)
• Pseudomonas aeruginosa is listed in the indication, but the label itself carries a footnote saying efficacy for that organism in the eye was studied in fewer than 10 infections. Evidence: unknown. (Source 3)
• The label states explicitly that the product is not indicated for this, and that safety and effectiveness below the age of 2 months have not been established. Any such use is outside the approved indication. Evidence: not-supported. (Source 4)

Interactions

• Other topical antimicrobial preparations and prolonged use generally (label): Using antibacterial drops for a long time can let organisms that the drug does not kill, including fungi, grow in their place. (Source 5)
• Folate and folic acid (trimethoprim is a folate antagonist) (theoretical): Trimethoprim works by blocking the enzyme that makes active folate in bacteria, and the label notes it may interfere with folic acid metabolism, which is why the label restricts use in pregnancy to cases where the benefit justifies the risk. The amount absorbed from eye drops is very small, and we found no study of a folate supplement interacting with the ophthalmic product. (Source 6)
• Contact lenses (label): The label advises that contact lenses should not be worn while there are signs and symptoms of an ocular bacterial infection. (Source 7)

Stopping it

• There is no dependence or withdrawal literature for this eye drop. The label's stopping instructions are about safety, not withdrawal: it directs that use be discontinued if a hypersensitivity reaction occurs, and that the drops are not for injection into the eye. (Source 4)
• The patient information section directs that use stop immediately and a physician be contacted if hypersensitivity reactions such as redness, irritation, swelling or pain persist or increase. (Source 7)
• The Cochrane review notes that antibiotic-treated participants were less likely to stop treatment early than placebo-treated ones, which is a measure of tolerability rather than of withdrawal. The review's own trial counts for this outcome ('13 trials, 5573 participants') and for persistent infection ('19 trials, 5280 participants') do not line up - fewer trials with more people - so neither participant total should be read as the size of the review. (Source 2)

What goes wrong

The local harms of the eye drop recorded in the label are irritation and hypersensitivity, and severe skin reactions and anaphylaxis have been reported. (Source 8)

• Official position, Low certainty.
• Size: not stated (spontaneous reports and label text)
• Who: People using the ophthalmic solution.
• How long: On instillation, within 48 hours, or at any time with extended use.
• Result: No rates are given. The label lists local irritation (redness, burning, stinging, itching) as most frequent, multiple reports of hypersensitivity with lid oedema, itching, redness, tearing or circumocular rash, and reports of Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis and anaphylaxis.
• Funding: regulatory label, not a funded study.

The most frequent adverse reaction to Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution is local irritation consisting of increased redness, burning, stinging, and/or itching. This may occur on instillation, within 48 hours, or at any time with extended use. There are also multiple reports of hypersensitivity reactions consisting of lid edema, itching, increased redness, tearing, and/or circumocular rash. Stevens- Johnson Syndrome/Toxic Epidermal Necrolysis (SJS/TEN) and anaphylaxis have been reported. Photosensitivity has been reported in patients taking oral trimethoprim.

In the Cochrane dataset, non-fluoroquinolone antibiotics - the class this combination belongs to - were associated with about four times the odds of treatment-related ocular complications compared with placebo, at very low certainty. (Source 9)

• Systematic review, Very low certainty.
• Size: 21 randomised trials, 8805 participants; 3 of the 21 tested non-fluoroquinolone antibiotics other than macrolides.
• Who: People with acute bacterial conjunctivitis.
• How long: Through the treatment period.
• Result: Non-fluoroquinolones RR 4.05 (95% CI 1.36 to 12.00) for treatment-associated ocular complications such as eye pain, discomfort and allergic reactions; fluoroquinolones RR 0.70 (95% CI 0.54 to 0.90). Certainty was very low.
• Funding: not stated in the abstract.

Limit of this finding: The Cochrane abstract and this authors' own summary of the same review describe the same two numbers in opposite directions: Cochrane writes that fluoroquinolones 'but not non-FQs ... may result in fewer participants with ocular side effects', while this summary writes that non-fluoroquinolones, but not fluoroquinolones, were likely to increase ocular complications. A reader should not take either sentence as settled; the figures themselves (RR 4.05 above 1 for non-fluoroquinolones, RR 0.70 below 1 for fluoroquinolones) are what both papers print, and the certainty is very low either way. Put plainly, so the direction is not read backwards: a risk ratio of 4.05 with a confidence interval of 1.36 to 12.00 lies entirely above 1, which means MORE ocular side effects on non-fluoroquinolones than on placebo, and non-fluoroquinolone is the class this eye drop belongs to. A risk ratio of 0.70 with an interval of 0.54 to 0.90 lies entirely below 1, which means FEWER side effects on fluoroquinolones. The Cochrane sentence's grammar reads as though both numbers were reductions; they are not.

Non-FQs (RR: 4.05; 95% CI: 1.36-12.00), but not FQs (RR: 0.70; 95% CI: 0.54-0.90), were likely to increase treatment-associated ocular complications such as eye pain, discomfort, and allergic reactions; the certainty of level of evidence was very low.

Resistance to antibiotics routinely used in eye practice remains high among bacteria cultured from the conjunctiva. (Source 10)

• Survey study, Low certainty.
• Size: 2214 conjunctival isolates collected 2009-2021.
• Who: Conjunctival isolates in a United States surveillance study (918 Staphylococcus aureus, 589 coagulase-negative staphylococci, 194 Streptococcus pneumoniae, 171 Pseudomonas aeruginosa, 342 Haemophilus influenzae)
• How long: 13 years of surveillance.
• Result: Staphylococci commonly resistant to azithromycin (>=54.8%) and oxacillin (>=29.3%); Streptococcus pneumoniae resistance notable for azithromycin (34.0%) and penicillin (28.9%); methicillin-resistant staphylococci showed multidrug resistance rates >=74.0%. Trimethoprim and polymyxin B are not among the drugs named in the abstract.
• Funding: not stated in the abstract; ARMOR is an industry-sponsored surveillance programme and a reviewer should confirm the funding statement in the paper.

Limit of this finding: This is in-vitro susceptibility surveillance, not a clinical outcome study, and the abstract does not report resistance to trimethoprim or polymyxin B specifically. It should not be read as a resistance rate for this product.

Staphylococci were commonly resistant to azithromycin (≥54.8%) and oxacillin (≥29.3%). Resistance among S. pneumoniae isolates was notable for azithromycin (34.0%) and penicillin (28.9%), while P. aeruginosa and H. influenzae isolates were highly susceptible to most tested antibiotics. Methicillin-resistant staphylococci demonstrated greater concurrent resistance to other antibiotics than methicillin-susceptible isolates and exhibited high rates of multidrug resistance

Systemic polymyxin B given by injection - a different product and route from the eye drop - carries a boxed warning for kidney and nerve toxicity. (Source 11)

• Official position, Certainty not rated.
• Size: not applicable (regulatory label)
• Who: Hospitalised patients given polymyxin B intramuscularly or intrathecally.
• How long: not applicable.
• Result: No rates are given in the boxed warning. It names albuminuria, cellular casts and azotemia as signs of nephrotoxicity, and irritability, weakness, drowsiness, ataxia, perioral paresthesia, numbness of the extremities and blurring of vision as neurotoxic reactions, and warns of respiratory paralysis from neuromuscular blockade.
• Funding: regulatory label, not a funded study.

Limit of this finding: This warning belongs to polymyxin B for injection, a systemic product. It is not a harm of the ophthalmic solution, which the label records produces peak serum polymyxin B of about 1 unit/mL after two drops.

RENAL FUNCTION SHOULD BE CAREFULLY DETERMINED AND PATIENTS WITH RENAL DAMAGE AND NITROGEN RETENTION SHOULD HAVE REDUCED DOSAGE. PATIENTS WITH NEPHROTOXICITY DUE TO POLYMYXIN B SULFATE USUALLY SHOW ALBUMINURIA, CELLULAR CASTS, AND AZOTEMIA.

In a hospital cohort of patients given intravenous polymyxin B - again a different product and route - acute kidney injury developed in 43% of them. (Source 12)

• Cohort study, Low certainty.
• Size: 84 patients.
• Who: Adults over 18 at a tertiary hospital in China who received polymyxin B for more than 48 hours; median age 80 years.
• How long: Median treatment duration 7.75 days; median time to acute kidney injury onset 5 days.
• Result: Acute kidney injury in 36 of 84 patients (42.9%) by KDIGO criteria; 11 (30.6%) of those discontinued polymyxin B and 8 required renal replacement therapy. Chronic kidney disease (OR 5.47, 95% CI 1.52-19.64, P = 0.01) and daily maintenance dose (OR 12.57, 95% CI 2.84-55.59) were independently associated with acute kidney injury.
• Funding: not stated in the abstract.

Limit of this finding: The abstract's sub-counts do not sum consistently: 11 of 36 is correctly given as 30.6%, but 8 of 36 would be 22.2%, not the 16.7% printed. A reader should not rely on the renal-replacement percentage. This is also an observational cohort in critically ill patients given intravenous polymyxin B, so it cannot establish that the drug caused the kidney injury, and it says nothing about the eye drop. The same abstract prints 'P = 0.00' for the maintenance-dose odds ratio, which is not a possible p-value and can only mean below 0.01. It also gives a median daily maintenance dose (1.82 mg/kg) higher than the median loading dose (1.73 mg/kg), the reverse of how polymyxin B is normally given, so nothing in this passage should be read as dosing. The indexed abstract also runs its section headings into the next word ('Patients and methodsThe study', 'ResultsThe majority'); that is the source's own text and has been left exactly as printed.

The development of AKI was observed in 36 patients (42.9%), with a median time to onset of 5 days (IQR: 3, 6). In the cohort of patients who developed AKI, 11 (30.6%) patients discontinued PMB therapy, while 8 (16.7%) patients required renal replacement therapy (RRT).

What the evidence supports

Across 21 placebo-controlled trials of topical antibiotics as a class for acute bacterial conjunctivitis, antibiotics improved clinical cure by about a quarter in relative terms, at moderate certainty. (Source 2)

• Systematic review, Moderate certainty.
• Size: 21 randomised trials, 8805 participants randomised; the clinical-cure estimate rests on 5 trials and 1474 participants.
• Who: People with acute bacterial conjunctivitis, in trials with heterogeneous eligibility criteria.
• How long: Outcome assessed at the end of therapy, with spontaneous resolution counted at days 4 to 9.
• Result: Clinical cure RR 1.26 (95% CI 1.09 to 1.46); 68.2% (504/739) cured on antibiotic versus 55.5% (408/735) on placebo, which is a difference of 12.7 percentage points, or about 8 people treated for one extra resolution (absolute figures computed by us from the proportions the review prints). Microbiological cure RR 1.53 (95% CI 1.34 to 1.74; 10 trials, 2827 participants). This is a class finding; polymyxin B with trimethoprim is not named in the abstract.
• Funding: not stated in the abstract; the review records research support including N.I.H. extramural funding.

Limit of this finding: Two things need stating plainly. This is a result for topical antibiotics as a class against placebo; polymyxin B with trimethoprim is not named anywhere in the abstract, so the 26% relative improvement is not a measured result for this eye drop. And the review's own trial counts do not line up: one outcome is reported from '13 trials, 5573 participants' and another from '19 trials, 5280 participants' - fewer trials with more people. Different outcomes can draw on different subsets of trials, but the two totals cannot both be straightforward subsets of one pool, so neither participant figure should be read as the size of the review.

Based on intention-to-treat (ITT) population, antibiotics likely improved clinical cure (resolution of clinical symptoms or signs) by 26% (RR 1.26, 95% CI 1.09 to 1.46; 5 trials, 1474 participants; moderate certainty) as compared with placebo. Subgroup analysis showed no differences by antibiotic class (P = 0.67) or treatment duration (P = 0.60). In the placebo group, 55.5% (408/735) of participants had spontaneous clinical resolution by days 4 to 9 versus 68.2% (504/739) of participants treated with an antibiotic.

A separate randomised trial in children found polymyxin B with trimethoprim no worse than moxifloxacin for clinical cure by the end of treatment. (Source 13)

• Randomized trial, Low certainty.
• Size: 124 patients enrolled; 114 completed the day 4-6 assessment and 89 the day 7-10 assessment.
• Who: Children aged 1 to 18 years with acute conjunctivitis; 80 of 124 (65%) had a recognised pathogen cultured.
• How long: 7 days of treatment, assessed at day 4-6 by telephone and day 7-10 by examination.
• Result: Parent-reported clinical cure at day 4-6: 42/58 (72%) on polymyxin B-trimethoprim versus 43/56 (77%) on moxifloxacin (noninferiority test P = .04). Clinical cure at day 7-10: 49/51 (96%) versus 37/39 (95%) (noninferiority test P <= .01). No significant difference in bacteriologic cure.
• Funding: not stated in the abstract.

Limit of this finding: The abstract reports 50/58 polymyxin B-trimethoprim patients as completing the day 7-10 evaluation but then gives the day 7-10 cure rate with a denominator of 51 (49/51). The two numbers do not agree, so a reader should not treat the day 7-10 denominator as settled. There is also no placebo arm, so a high cure rate in both arms cannot be separated from spontaneous resolution. The abstract's pathogen sub-counts do not sum either: it reports 80 patients (65%) with recognised pathogens and then lists 55 Haemophilus influenzae, 22 Streptococcus pneumoniae and 4 Moraxella catarrhalis, which comes to 81. One patient carrying two organisms would explain it, but the abstract does not say so.

One hundred fourteen (56/62 moxifloxacin and 58/62 polymyxin B-trimethoprim) completed the 4-6 day evaluation, with 43/56 (77%) of the moxifloxacin group and 42/58 (72%) of the polymyxin B-trimethoprim group clinically cured according to parents (noninferiority test P = .04).

What the evidence does not support

More than half of people given placebo for acute bacterial conjunctivitis recovered without any antibiotic, and the review states that many cases are self-limited. (Source 14)

• Systematic review, Moderate certainty.
• Size: Placebo arms of 5 trials, 735 participants for the clinical-cure comparison.
• Who: People with acute bacterial conjunctivitis.
• How long: Days 4 to 9.
• Result: 55.5% (408/735) of placebo-treated participants had spontaneous clinical resolution by days 4 to 9.
• Funding: not stated in the abstract.

Antibiotics are generally prescribed on the basis that they may speed recovery, reduce persistence, and prevent keratitis. However, many cases of acute bacterial conjunctivitis are self-limited, resolving without antibiotic therapy.

In the only modern randomised comparison that measured speed of recovery, polymyxin B with trimethoprim cleared the signs and symptoms of bacterial conjunctivitis in far fewer children by 48 hours than moxifloxacin did. (Source 15)

• Randomized trial, Low certainty.
• Size: 84 eyes of 56 patients.
• Who: Patients younger than 18 years with a clinical diagnosis of bacterial conjunctivitis, multicentre.
• How long: Assessments at 24 and 48 hours, with a day-7 telephone interview.
• Result: Complete resolution of ocular signs and symptoms at 48 hours in 44% of the polymyxin/trimethoprim group versus 81% of the moxifloxacin group (P = .001). No adverse events were reported. Single comparison, no placebo arm, and the trial is small.
• Funding: not stated in the abstract; two of the named authors are affiliated to the manufacturer of the comparator in the published author list, which a reviewer should check against the paper itself.

Limit of this finding: This trial compared two active drugs at different dosing frequencies (four times daily versus three times daily) and had no placebo arm, so it cannot show whether either drug beats no treatment. The outcome was speed of resolution at 48 hours, not final cure. The paper's conclusion also goes beyond what it measured: it states that moxifloxacin 'is cost-effective' and that it reduces 'disease transmission', but neither cost nor transmission was an endpoint - the methods describe resolution of ocular signs and symptoms, cultures, patient-rated symptoms and adverse events. The indexed record lists four authors and prints an academic affiliation for the first only (Granet DB); the other three, two of whom (Stroman D, Cockrum P) independent checking identifies as employees of the favoured product's manufacturer, carry no affiliation in the record, and the abstract carries no funding statement at all - the record is flagged only 'Research Support, Non-U.S. Gov't'. So the cost and transmission claims should be read as an author team's opinion, with a probable industry interest that the published abstract does not disclose, rather than as a finding.

At the 48-hour visit, complete resolution of ocular signs and symptoms was observed in 81% of the patients treated with moxifloxacin and 44% of the patients treated with polymyxin/trimethoprim (P = .001). No adverse events were reported.

Where the research disagrees

Whether non-fluoroquinolone topical antibiotics (the class that includes this combination) increase or decrease ocular side effects

• The 2023 Cochrane review's own abstract, systematic-review: When compared with placebo, FQs (RR 0.70, 95% CI 0.54 to 0.90) but not non-FQs (RR 4.05, 95% CI 1.36 to 12.00) may result in fewer participants with ocular side effects. However, the estimated effects were of very low certainty. (Source 2)
• The same review team's summary of the same review in the American Journal of Ophthalmology, systematic-review: Non-FQs (RR: 4.05; 95% CI: 1.36-12.00), but not FQs (RR: 0.70; 95% CI: 0.54-0.90), were likely to increase treatment-associated ocular complications such as eye pain, discomfort, and allergic reactions; the certainty of level of evidence was very low. (Source 9)

Whether this combination is as good as a fluoroquinolone

• Williams and colleagues, 2013, in a single-blinded randomised trial in children, rct: Clinical cure was observed in 37/39 (95%) of the moxifloxacin and 49/51 (96%) of the polymyxin B-trimethoprim treated groups (noninferiority test P ≤ .01). Clinical cure rates for culture positive and negative conjunctivitis were not different. There was no statistically significant difference in bacteriologic cure rates between the 2 groups. (Source 13)
• Granet and colleagues, 2008, in a multicentre randomised comparison of speed of effect, rct: Moxifloxacin 0.5% administered three times daily is safe and cures bacterial conjunctivitis more effectively and significantly faster than polymyxin/trimethoprim dosed four times daily. The majority of patients were cured and symptom-free by 48 hours. (Source 16)

How much

• Reference intake: There is no reference intake for a prescription antibiotic. Dosing is set by the prescriber. As a position, the label of the ophthalmic solution records, with an effective date of 2025-09-05, that in mild to moderate infections one drop is instilled in the affected eye every three hours to a maximum of 6 doses per day for 7 to 10 days. (Source 17)
• Upper limit: The label's stated ceiling, as a position dated 2025-09-05, is a maximum of 6 doses per day over a 7 to 10 day course. No upper intake level in the nutritional sense exists for this product. (Source 17)
• Studied: The 2008 multicentre comparison randomised patients to one drop of polymyxin B sulfate/trimethoprim four times daily for 7 days or one drop of 0.5% moxifloxacin three times daily for 7 days. (Source 18)
• Studied: The pharmacology study recorded in the label instilled 2 drops of ophthalmic solution containing 1 mg trimethoprim and 10,000 units polymyxin B per mL in 11 volunteers. (Source 1)

A common belief, and what the research shows

The belief: Pink eye needs antibiotic drops or it will not clear up.

What the research shows: The 2023 Cochrane review of 21 placebo-controlled trials recorded that most cases are self-limited: However, many cases of acute bacterial conjunctivitis are self-limited, resolving without antibiotic therapy. In the trials it pooled, 55.5% of placebo-treated participants resolved by days 4 to 9 against 68.2% on an antibiotic, so the antibiotic adds to a recovery that mostly happens anyway, and the review's own authors describe the gain as modest: The findings of this update suggest that the use of topical antibiotics is associated with a modestly improved chance of resolution in comparison to the use of placebo.

Questions and answers

What is it?

It is an eye drop containing two antibacterial drugs, trimethoprim and polymyxin B sulfate. Trimethoprim is a synthetic antibacterial that blocks a bacterial folate enzyme; polymyxin B is a cyclic lipopeptide antibiotic that kills Gram-negative bacteria. The two are combined so that the pair covers more of the bacteria that cause eye-surface infections than either would alone. (Source 1)

What does it do in the body?

On the surface of the eye, trimethoprim stops susceptible bacteria from making the folate they need to build DNA and proteins, and polymyxin B punches holes in the outer membrane of Gram-negative bacteria. Very little gets into the bloodstream: in 11 volunteers given two drops, peak serum levels were about 0.03 mcg/mL of trimethoprim and 1 unit/mL of polymyxin B. (Source 1)

Is it good or bad for you?

For bacterial conjunctivitis the honest answer is a small benefit on top of a condition that usually settles by itself. In the 2023 Cochrane review of topical antibiotics as a class, 68.2% treated versus 55.5% of placebo-treated participants had resolved by days 4 to 9. Against that, the class of antibiotics this drug belongs to was linked to more ocular side effects than placebo at very low certainty, and resistance among eye bacteria is a running concern. (Source 19)

How do you get more of it?

It is a prescription-only medicine and there is no food, supplement or behaviour that produces it. The only way anyone receives it is a prescription, and the label states as a position that in mild to moderate infections one drop goes into the affected eye every three hours, to a maximum of six doses a day, for seven to ten days. That is the label's position on dosing, not advice to a reader. (Source 17)

If it is harmful, what reduces it?

The drug is cleared by stopping the drops; there is no literature on tapering it. The label instructs that use be discontinued if a hypersensitivity reaction occurs, and the patient information section says to stop immediately and contact a physician if redness, irritation, swelling or pain persist or increase. (Source 7)

Why might someone be low in it or missing it?

This question does not apply in the usual sense: nobody is naturally low in a manufactured antibiotic. The reasons someone would not be given it are that they are hypersensitive to a component, which is the label's only contraindication, or that they are under two months old, where safety and effectiveness have not been established, or that their infection resolved without it. (Source 20)

Which whole foods contain it or feed it?

No whole food contains polymyxin B or trimethoprim. Trimethoprim is described in the label as a synthetic antibacterial drug and polymyxin B as a cyclic lipopeptide antibiotic; neither is a dietary constituent, and food has no role in supplying them. (Source 1)

What happens if you do not have it?

Not using the drops is not a deficiency state. For acute bacterial conjunctivitis the Cochrane review found that most untreated cases resolve on their own, with 55.5% of placebo-treated participants clear by days 4 to 9, and that antibiotics were also linked to a lower chance of persistent clinical infection (RR 0.73). The label's stated rationale for treating is to speed recovery, reduce persistence and prevent keratitis. Two cautions on those numbers: they are for topical antibiotics as a class rather than for this eye drop specifically, and the review's own trial counts for the two outcomes do not line up - 13 trials with 5,573 people for treatment incompletion against 19 trials with 5,280 people for persistent infection. (Source 2)

How can you test for it?

There is no blood test for the drug in ordinary care. The test that matters for the condition is a conjunctival culture, which the trials used to confirm infection and to judge microbiological cure, and it is imperfect: in the 2013 children's trial only 65% of those with clinically diagnosed acute conjunctivitis had a recognised pathogen cultured, and clinical cure rates did not differ between culture-positive and culture-negative children. One caution on that 65%: the trial's own pathogen sub-counts do not sum - it reports 80 of 124 patients with a recognised pathogen and then lists 55, 22 and 4 isolates, which comes to 81 - so treat 65% as approximate. (Source 13)

References

1. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: CLINICAL PHARMACOLOGY. 2025. Read the source
2. The Cochrane database of systematic reviews. Antibiotics versus placebo for acute bacterial conjunctivitis.. 2023. PMID 36912752, DOI 10.1002/14651858.CD001211.pub4. Read the source
3. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: INDICATIONS AND USAGE. 2025. Read the source
4. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: WARNINGS. 2025. Read the source
5. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: General. 2025. Read the source
6. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: Teratogenic Effects. 2025. Read the source
7. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: Information for Patients. 2025. Read the source
8. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: ADVERSE REACTIONS. 2025. Read the source
9. American journal of ophthalmology. Antibiotics Versus Placebo for Acute Bacterial Conjunctivitis: Findings From a Cochrane Systematic Review.. 2024. PMID 37482371, DOI 10.1016/j.ajo.2023.06.027. Read the source
10. Diagnostic microbiology and infectious disease. Antibiotic resistance of bacterial pathogens isolated from the conjunctiva in the Antibiotic Resistance Monitoring in Ocular micRoorganisms (ARMOR) surveillance study (2009-2021).. 2024. PMID 37918187, DOI 10.1016/j.diagmicrobio.2023.116069. Read the source
11. DailyMed (U.S. National Library of Medicine), SPL setid b56f18c0-ef5e-4ed9-a5af-f79f3cd189b6. Polymyxin B for Injection, USP 500,000 Units/vial (Xellia Pharmaceuticals USA LLC) - DailyMed Structured Product Label, section: WARNING. 2026. Read the source
12. Science progress. The association between acute kidney injury and Polymyxin B treatment: A retrospective cohort study.. 2025. PMID 40965976, DOI 10.1177/00368504251379362. Read the source
13. The Journal of pediatrics. A single-blinded randomized clinical trial comparing polymyxin B-trimethoprim and moxifloxacin for treatment of acute conjunctivitis in children.. 2013. PMID 23092529, DOI 10.1016/j.jpeds.2012.09.013. Read the source
14. The Cochrane database of systematic reviews. Antibiotics versus placebo for acute bacterial conjunctivitis.. 2023. PMID 36912752, DOI 10.1002/14651858.CD001211.pub4. Read the source
15. Journal of pediatric ophthalmology and strabismus. A multicenter comparison of polymyxin B sulfate/trimethoprim ophthalmic solution and moxifloxacin in the speed of clinical efficacy for the treatment of bacterial conjunctivitis.. 2008. PMID 19043945, DOI 10.3928/01913913-20081101-07. Read the source
16. Journal of pediatric ophthalmology and strabismus. A multicenter comparison of polymyxin B sulfate/trimethoprim ophthalmic solution and moxifloxacin in the speed of clinical efficacy for the treatment of bacterial conjunctivitis.. 2008. PMID 19043945, DOI 10.3928/01913913-20081101-07. Read the source
17. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: DOSAGE AND ADMINISTRATION. 2025. Read the source
18. Journal of pediatric ophthalmology and strabismus. A multicenter comparison of polymyxin B sulfate/trimethoprim ophthalmic solution and moxifloxacin in the speed of clinical efficacy for the treatment of bacterial conjunctivitis.. 2008. PMID 19043945, DOI 10.3928/01913913-20081101-07. Read the source
19. The Cochrane database of systematic reviews. Antibiotics versus placebo for acute bacterial conjunctivitis.. 2023. PMID 36912752, DOI 10.1002/14651858.CD001211.pub4. Read the source
20. DailyMed (U.S. National Library of Medicine), SPL setid 71900075-5288-4f4b-afe2-098436fa2173. Polymyxin B Sulfate and Trimethoprim Ophthalmic Solution, USP (Sandoz Inc) - DailyMed Structured Product Label, section: CONTRAINDICATIONS. 2025. Read the source