For healthcare professionals · clinical synopsis

Bacterial keratitis and corneal ulcer

A practical clinical guide to severity assessment, microbiological diagnosis, empirical antibacterial treatment, response monitoring and timely revision of management in bacterial keratitis and corneal ulcer.

ICD-10: H16.0 — corneal ulcer; final coding depends on the clinical phenotypeBased on: Russian Ministry of Health “Keratitis” guideline 2025 + AAO PPP 2024Reviewed: 20 Jul 2026
Corneal infiltrateClinical photograph from the author’s archive, M. Y. Shemyakin.
Clinical sequence: assess severity → obtain specimens → start treatment → monitor the response

For clinical practice in the Russian Federation, current Ministry of Health clinical guidelines, official prescribing information and local standard operating procedures remain the primary regulatory references. International guidance is presented separately as an additional clinical framework.

1–2–3risk criteria
24 hreview in severe disease
48 hreassess diagnosis and treatment
i
Author-prepared clinical synopsis; not an official guideline text

This material is intended for ophthalmologists and other qualified healthcare professionals. The Russian section is based on the Ministry of Health clinical guideline “Keratitis” 2025 (ID 1010_1). The international section is supplementary and does not replace national guidance, official prescribing information or local SOPs.

01 · URGENCY

Identify potentially vision-threatening disease

The 1–2–3 ACT criteria are used here as an additional risk-stratification tool. Meeting any criterion should increase concern for vision-threatening microbial keratitis and supports microbiological sampling and consideration of more intensive initial treatment.

Anterior chamber cell ≥ 1+

Anterior chamber cell ≥1+ by SUN grading: 6–15 cells in a 1 × 1 mm slit beam field.

Infiltrate ≥ 2 mm

Additional high-risk features include multiple or satellite infiltrates, deep stromal involvement and progressive stromal melt.

≤ 3 mm from centre

The nearest edge of the stromal infiltrate lies no more than 3 mm from the anatomical centre of the cornea.

If any 1–2–3 ACT criterion is met:

treat the case as potentially vision-threatening → obtain corneal scraping for microscopy and culture before the first antibiotic dose when this does not cause a clinically meaningful treatment delay → consider a more intensive initial regimen, including locally prepared fortified antibiotics where appropriate, with daily review until stabilisation in severe disease.

!
Urgent corneal-specialist review and assessment for admission

Progressive thinning, descemetocele or perforation; rapidly increasing infiltrate area or depth; extension to sclera or intraocular structures; infiltrate involving a surgical wound, corneal suture or lamellar interface; suspected gonococcal infection; or inability to deliver the required treatment frequency and follow-up.

02 · INITIAL ASSESSMENT

Document the baseline before treatment

Recording the same parameters at each review makes change in the epithelial defect, stromal infiltrate and inflammatory response objectively comparable.

HistoryRisk factors
  • Contact-lens wear, overnight wear, exposure of lenses to tap water, swimming pools or hot tubs
  • Trauma, particularly with vegetative material
  • Recent surgery, corneal sutures or keratoplasty
  • Previous topical antibiotics or corticosteroids
  • Ocular-surface disease, immunocompromise or diabetes mellitus
Visual function and IOPBaseline measurements
  • Uncorrected and best-corrected visual acuity in each eye
  • Pain severity, photophobia and character of ocular discharge
  • Intraocular pressure only when there is no major thinning or risk of perforation
  • Status of the fellow eye
Slit-lamp examinationLesion morphology
  • Location, size and depth of the infiltrate
  • Size of the epithelial defect
  • Margins, satellite lesions, stromal melt and degree of thinning
  • Anterior chamber cell, fibrin, hypopyon height and endothelial plaque
  • Anterior-segment photography using a comparable scale
SMALL PERIPHERAL LESIONPeripheral infiltrate approximately 1 mm

For a small peripheral infiltrate, document its distance from the corneal centre, stromal depth, epithelial-defect size and subsequent change.

Clinical archive of M. Y. Shemyakin

PARACENTRAL ZONECorneal ulcer approximately 2 mm

Paracentral location and lesion size increase the need for microbiological diagnosis, adequate initial treatment intensity and early reassessment.

Clinical archive of M. Y. Shemyakin

Suggested examination template

Stromal infiltrate: ___ × ___ mm; location: ___; distance from nearest infiltrate edge to corneal centre: ___ mm; stromal depth: ___; epithelial defect: ___ × ___ mm; corneal thinning: ___%; anterior chamber cell by SUN grade: ___; hypopyon: ___ mm; scaled photography: performed / not performed.

03 · MICROBIOLOGY

When to obtain corneal specimens

A small peripheral infiltrate without stromal melt or marked anterior chamber inflammation may sometimes be treated empirically. Central, large, deep, postoperative or atypical disease warrants microbiological sampling before antimicrobial treatment when this does not delay therapy.

  • Central infiltrate; lesion >2 mm; two or more adjacent lesions; marked deep stromal involvement or stromal melt.
  • Anterior chamber cell ≥1+ by SUN grading, hypopyon or other features of vision-threatening disease.
  • Previous corneal surgery; infiltrate involving a suture, surgical wound or lamellar interface.
  • Chronic or recurrent course, multiple infiltrates, atypical morphology or failure to respond as expected to broad-spectrum antibacterial therapy.
  • Suspicion of fungal, Acanthamoeba, Nocardia or nontuberculous mycobacterial infection.
Sampling techniqueSample the active edge and infiltrate base

Corneal scraping is obtained with a sterile instrument after topical anaesthesia. Superficial conjunctival discharge alone is less informative. An infected or loose corneal suture should be removed and sent for microbiological examination when clinically indicated.

Additional specimensContact lens, case and storage solution

In contact-lens wearers, culture or analysis of the lens, storage case and solution may provide additional information. A hypopyon in isolated keratitis is often sterile; anterior-chamber or vitreous sampling is not routine and is reserved for concern about intraocular extension.

Clinical situationCultureMinimum diagnostic setWhen to expand
Small peripheral infiltrateNo stromal melt or substantial anterior chamber inflammationOptionalClinical assessment and close follow-upIf worsening occurs, obtain microbiology before changing therapy when feasible
Central / large / deep lesionVision-threatening featuresYesDirect microscopy including Gram stain; culture on blood and chocolate agar and in liquid mediaAdd Sabouraud and other media according to history and suspected etiology
Atypical or postoperative diseaseYesBacterial culture plus targeted stainsKOH and/or calcofluor white, acid-fast stains, special media for Acanthamoeba and nontuberculous mycobacteria as indicated
04 · VISUAL DIFFERENTIAL DIAGNOSIS

Use morphology to refine — not replace — etiologic testing

Clinical appearance can help broaden or narrow the differential diagnosis but cannot reliably identify the organism by itself. Images are presented as attributed clinical examples.

AUTHOR’S CLINICAL ARCHIVE

The photographs illustrate clinical morphology and do not replace microbiological or, when indicated, molecular confirmation.

SELECTED CLINICAL PATTERN

Bacterial keratitis

Acute onset, a suppurative stromal infiltrate, an epithelial defect and anterior chamber inflammation are typical. Morphology alone does not identify the causative organism.

Features that increase probability
  • Contact-lens wear, corneal trauma, surgery or corneal sutures
  • Purulent discharge and rapidly progressive stromal disease
  • Hypopyon, deep infiltration and stromal melt
What to do next
  • Assess for vision-threatening features
  • Perform corneal scraping in severe or atypical disease
  • Start antibacterial therapy immediately after sampling when sampling is indicated
!
A clinical photograph does not establish microbiological etiology

Visual assessment should trigger appropriate expansion of the differential diagnosis and investigations; it should not create false confidence about the causative organism.

05 · TREATMENT

Russian clinical guidance, international guidance and practical differences

The sources are deliberately separated so that national Russian recommendations are not conflated with supplementary international approaches.

Russian practice
Primary framework for use in the Russian Federation

Russian Ministry of Health clinical guideline “Keratitis” 2025, official prescribing information for the specific medicinal product and local SOPs. International regimens are shown separately.

Russian Federation · Ministry of Health guideline 2025

Clinical guideline “Keratitis”

ID 1010_1 · adults and children · scheduled review no later than 2027

Topical antibacterial therapy
  • Ofloxacin: 1–2 drops every 4–6 hours.
  • Ciprofloxacin: 1–2 drops every 4 hours.
  • Levofloxacin: 1–2 drops every 2 hours, not exceeding 8 instillations per day.
  • Netilmicin: 1–2 drops 3 times daily; tobramycin: 1–2 drops every 4 hours.
  • Treatment duration is determined by the approved prescribing information and clinical response; the guideline gives an indicative course of 5–10 days.
Additional provisions
  • The guideline also includes topical antiseptic treatment in accordance with the relevant prescribing information.
  • When keratitis is complicated by iridocyclitis and there is a risk of posterior synechiae, mydriatic/cycloplegic treatment may be considered in accordance with the guideline and approved product information.
  • Systemic antibacterial treatment may be prescribed when clinically indicated, including extensive or complicated bacterial disease.
RECOMMENDATION PROFILE
SourceRussian Ministry of Health “Keratitis” guideline 2025
StatusPrimary national reference in the Russian Federation
Evidence gradingStrength of recommendation C; certainty of evidence 5 for topical antibacterial treatment
Verified19 Jul 2026
Document limitation: the guideline lists drugs and dosing schedules but stratifies initial treatment intensity by infiltrate size, depth and central location less explicitly than AAO PPP and the 1–2–3 ACT framework.
06 · RESPONSE MONITORING

Judge treatment by reproducible clinical change

In severe disease, examine the patient daily until stabilisation. Failure to improve or progression during the first 24–48 hours requires reassessment of the diagnosis, treatment delivery, organism and need for surgical intervention.

Pain and discharge decreaseSubjective improvement should be consistent with lesion morphology.
Infiltrate borders become more definedThere is no increase in infiltrate area or depth and no progression of stromal melt.
Stromal density and surrounding oedema decreaseThe epithelial defect and thinning are not increasing.
Anterior chamber inflammation decreasesAnterior chamber cell, fibrin and hypopyon should trend down.
Possible early exceptionPseudomonas aeruginosa and other Gram-negative keratitis

Inflammatory signs can transiently appear worse during the first 24–48 hours even when the organism is susceptible. Treatment changes should therefore be based on the overall pattern, particularly infiltrate area and depth, epithelial-defect size and stromal destruction.

When to reconsider therapyNo improvement or stabilisation within 24–48 hours

Check the actual instillation frequency, microscopy and culture results, antimicrobial susceptibility, medication-related epithelial toxicity, atypical or mixed infection, biofilm, infected sutures or foreign material, and whether surgical treatment is indicated.

OUTCOMECorneal opacity after keratitis

Even after active infection has resolved, visual outcome is determined by the location and density of residual scarring.

Clinical archive of M. Y. Shemyakin

RISK OF STROMAL LOSSThinning in a penetrating corneal graft

This pattern requires differentiation between persistent infection and non-infectious stromal thinning, together with timely assessment of perforation risk.

Clinical archive of M. Y. Shemyakin

Standardise every follow-up examinationInfiltrate size and depth · epithelial defect · degree of thinning · anterior chamber cell · hypopyon height · UCVA/BCVA · photography at the same scale
07 · NO IMPROVEMENT AFTER 24–48 HOURS

Reassess the cause of treatment failure systematically

This is a structured reassessment framework, not an automated prescribing system. It is intended to prevent escalation of the same treatment without reconsidering the diagnosis.

PRIMARY REASSESSMENT HYPOTHESIS

The presumed etiology may be wrong

Reassess the possibility of fungal, Acanthamoeba, herpetic, Nocardia or nontuberculous mycobacterial keratitis, particularly when the history or morphology is atypical.

  1. Reassess the history, risk factors and clinical morphology.
  2. Expand microscopy, culture and molecular testing when indicated.
  3. Until the etiology is clarified, avoid treatment that could worsen an alternative infection, especially premature topical corticosteroids.
NON-INFECTIOUS KERATOPATHYNeurotrophic keratitis

A persistent epithelial defect and stromal damage may reflect impaired corneal innervation and epithelial repair rather than persistent infection.

Clinical archive of M. Y. Shemyakin

Clinical implicationDo not simply increase antibacterial intensity without reassessment

When response is poor, distinguish persistent infection from medication-related epithelial toxicity and neurotrophic disease by integrating changes in the infiltrate, epithelial defect, stromal destruction and microbiology.

08 · CLINICAL CASE

Choose the most appropriate next step

Composite educational scenario. The photograph illustrates the morphology of a paracentral corneal ulcer and does not depict the described patient.

ILLUSTRATIVE CLINICAL PHOTOGRAPHparacentral ulcer approximately 2 mmClinical archive of M. Y. Shemyakin · image is not from the described patient
SCENARIO

A 43-year-old contact-lens wearer presents with pain and reduced vision

There is a 2.5-mm central stromal infiltrate with a corresponding epithelial defect, anterior chamber cell 1+ by SUN grading and a 1-mm hypopyon. No antibacterial therapy has yet been used.

What is the most appropriate first step?
Select an option to reveal the clinical rationale.
09 · PRIMARY SOURCES

Russian and international source documents

Primary sources open in a new tab. This synopsis should be medically re-reviewed when a core guideline or prescribing source is updated.

2025
Russian Ministry of Health. Clinical guideline “Keratitis” National guidelineID 1010_1. Approved by the Scientific and Practical Council of the Ministry of Health of the Russian Federation. Adults and children; scheduled revision no later than 2027.
Guideline registry ↗
2024
American Academy of Ophthalmology. Bacterial Keratitis Preferred Practice Pattern®Ophthalmology. 2024;131(4):P87–P133. DOI: 10.1016/j.ophtha.2023.12.035.
Open ↗
2020
Ung L, et al. Validation of a Comprehensive Clinical Algorithm for the Assessment and Treatment of Microbial KeratitisAmerican Journal of Ophthalmology. 2020;214:97–109. Source of the 1–2–3 ACT framework.
PubMed ↗
2025
Shemyakin MY, Chernakov AS. Contemporary approach to rational treatment of bacterial keratitis before penetrating keratoplastyClinical Ophthalmology. 2025;25(3):208–214. DOI: 10.32364/2311-7729-2025-25-3-8. Original publication in Russian.
Article ↗
2023
Arzhimatova GSh, et al. Management of patients with bacterial corneal ulcers in a multidisciplinary hospitalOphthalmology in Russia. 2023;20(3):572–579. DOI: 10.18008/1816-5095-2023-3-572-579. CC BY 4.0. Original publication in Russian.
Article ↗

Keep this material in your professional workspace on this device.

New and updated clinician materials are announced in the Telegram channel.

Get clinician updates