Four in ten hospitalised mpox patients developed eye disease, a Congolese cohort finds

Eye disease has been a recognised complication of mpox for as long as the disease has been described, but it has rarely been measured properly during an outbreak — which requires an ophthalmologist, repeatedly, at the bedside. A prospective cohort published on 27 August did that in South Kivu, in the eastern Democratic Republic of the Congo, and found ocular involvement in 43% of hospitalised patients with laboratory-confirmed clade Ib mpox [s1].

What was done

MBOTE-EYE was a substudy nested inside a prospective clinical cohort. All hospitalised patients with PCR-confirmed mpox within the previous 48 hours were eligible. Participants underwent ophthalmological examination at enrolment, at discharge, and again on days 29 and 59, and conjunctival swabs were collected for monkeypox virus PCR [s1].

Between 28 October 2024 and 30 June 2025, 310 participants were enrolled. Median age was 14.0 years (IQR 2.5–25.0) and 166 (54%) were female [s1]. That age distribution is itself part of the story: this is not a cohort of adults.

What was found

At enrolment, conjunctival disease was present in 112 participants (36%, 95% CI 31–42), corneal disease in 24 (8%, 5–11), and anterior uveitis in two (1%, 0–2) [s1]. Over follow-up, 134 participants (43%, 38–49) developed mpox-related ophthalmic disease, most often bilaterally [s1]. Blindness following ulcerative keratitis occurred in two of 224 participants (1%, 0–3) [s1].

Two risk factors stood out in the mixed-effects Poisson models. Periorbital lesions carried an adjusted risk ratio of 3.65 (95% CI 2.42–5.49) and severe acute malnutrition an adjusted risk ratio of 3.16 (1.22–8.22) [s1]. Among the 89 patients with conjunctival PCR results, cycle-threshold values below 25 — indicating higher viral loads — occurred exclusively during active ocular disease [s1].

The authors' interpretation is that ophthalmic involvement in clade Ib mpox is common and frequently bilateral, carries a substantial burden of keratitis and a risk of vision loss, and that systematic eye examinations should be integrated into mpox care while targeted ophthalmic therapies are developed [s1].

How it squares with the earlier South Kivu data

A cross-sectional study published in February in PLOS Global Public Health examined 366 hospitalised patients with laboratory-confirmed mpox across four South Kivu health zones between 1 November 2024 and 31 January 2025 [s2]. That cohort was younger still: median age 8 years (IQR 4–16), with 71% under 15 and 45.8% under five [s2]. Ocular symptoms or disease were recorded in 190 patients (51.9%); conjunctivitis in 113 (30.9%); blepharoconjunctivitis in 7.4%; blepharitis in 5.2%; keratoconjunctivitis in 3.3%; ulcerative keratitis in 2.5% and non-ulcerative keratitis in 2.2% [s2]. Visual impairment was recorded in 5.7% and blindness in 2.5% [s2].

The two studies do not report identical numbers, and should not be expected to: one is a cross-sectional snapshot with a single routine examination, the other a longitudinal cohort with four scheduled assessments and a stricter case definition of ophthalmic disease. The cross-sectional study recorded a higher proportion with any ocular symptom or finding at a single point; the prospective cohort recorded a higher proportion developing defined ophthalmic disease across two months. Both point the same way: eye involvement in this outbreak is not an occasional complication.

The blindness figures differ more than is comfortable — 2.5% of 366 in the earlier study, roughly 1% of 224 in the newer one [s1][s2]. Different denominators, different ascertainment and different follow-up windows can all produce that gap, and neither study is designed to arbitrate it. The honest summary is that severe vision loss occurred in both cohorts at a low single-digit percentage, with the point estimate uncertain.

Why the malnutrition signal matters

The adjusted risk ratio of 3.16 for severe acute malnutrition sits at the join between two crises [s1]. Eastern DRC's mpox outbreak is unfolding in a population where acute malnutrition is widespread, and the finding implies that nutritional status is not merely a background variable but a modifier of who loses vision. It also suggests that an intervention with no ophthalmic component — feeding — may bear on ophthalmic outcomes. This is an observational association within a single cohort, not a demonstration that correcting malnutrition prevents keratitis.

What the study cannot establish

Both studies enrolled hospitalised patients, so the estimates describe severity among people sick enough to be admitted, not among everyone infected. Neither was a trial: there was no randomised comparison of ophthalmic treatment strategies, and the prospective cohort describes the clinical course "under supportive management" [s1]. Loss to follow-up between day 29 and day 59 is a standard limitation of outbreak cohorts. And the findings are specific to clade Ib in this region.

What to watch

The concrete follow-on the authors name is a targeted ophthalmic therapy tested in a trial [s1]. In the meantime, the operational question is whether eye examination becomes routine in mpox treatment centres, which requires ophthalmic staff in places that generally do not have them.

Sources

• Prevalence, characteristics, and clinical course of mpox-related ophthalmic disease (MBOTE-EYE): a prospective cohort study in South Kivu, Democratic Republic of the Congo, The Lancet Global Health, 27 August 2026
• Ocular manifestations of Clade Ib mpox in four South-Kivu health zones, Democratic Republic of the Congo, PLOS Global Public Health, 18 February 2026