Design: Population-based retrospective register cohort
Population: 979,534 cataract operations in Sweden, 2018–2024
Primary outcome: Postoperative endophthalmitis
Effect: Overall 0.013%; cefuroxime 0.017%, cefuroxime plus ampicillin 0.011%, moxifloxacin 0.010%; no significant difference

The Swedish National Cataract Register covered 979,534 cataract operations from 2018 to 2024. There were 131 cases of endophthalmitis, an incidence of 0.013%, among the lowest reported anywhere.

Rates were 0.017% with intracameral cefuroxime alone, 0.011% with cefuroxime plus ampicillin and 0.010% with moxifloxacin. No regimen was significantly better. Enterococci, which cefuroxime does not cover, were commoner with cefuroxime alone, and eyes infected despite moxifloxacin more often recovered to 20/40. Male sex, ocular comorbidity and capsule complications were associated with higher odds; immediate sequential bilateral surgery was associated with lower odds, probably reflecting case selection.

The headline is the low baseline rate with near-universal intracameral prophylaxis. Differences between agents were small and not statistically clear.

For Indian practice, where intracameral antibiotic use is still not universal and moxifloxacin is widely available, the message is to use an intracameral agent in every case. Moxifloxacin is a reasonable choice given its enterococcal cover. Posterior capsule rupture should prompt extra vigilance.

Use an intracameral antibiotic in every cataract operation
Moxifloxacin and cefuroxime-based regimens performed similarly here
Be more vigilant after capsule complications, which carried higher endophthalmitis odds
Immediate sequential bilateral surgery was not linked to higher endophthalmitis risk
Counsel patients to report pain or falling vision after surgery at once

Why it matters

It shows the infection rate achievable when intracameral prophylaxis is routine.

The statistics, in plain English

With only 131 infections in nearly a million operations, the differences between 0.010% and 0.017% are too small to separate from chance. The associations with patient factors are observational and may reflect who was chosen for which surgery.