Liquid versus solid culture media for the microbiological diagnosis of eye infections: a retrospective comparative...

Ocular infections are an important cause of preventable vision loss, yet culture yield is limited by small specimen volumes and low bacterial loads. Comparisons of liquid and solid media in ocular infection are scarce. We evaluated four bedside-inoculated media: two liquid (Brucella broth, thioglycolate broth) and two solid (chocolate agar, 5% sheep blood agar).

We retrospectively analysed 102 ocular specimens collected between January 2023 and August 2025, each inoculated simultaneously onto all four media at the bedside. Because solid media were incubated for 7 days and liquid media for up to 21 days, positivity was compared at matched cut-offs of 24 hours, 48 hours and 7 days using McNemar and Cochran's Q tests, and all comparisons were repeated after excluding contaminant cultures.

Growth was detected in 77.5% of specimens. At least one liquid medium was positive in 72.5% of cases versus 51.0% for solid media (p=0.0002). Brucella broth had the highest positivity (65.7%), followed by thioglycolate broth (60.8%); both solid media yielded 49.0%. Twenty-nine isolates, including Serratia marcescens, Streptococcus pneumoniae and Candida parapsilosis, were recovered exclusively from liquid media. At the matched 7-day cut-off, liquid media remained superior before (70.6% vs 51.0%; p=0.0005) and after (69.8% vs 53.1%; p=0.0022) excluding contaminants, with an identical odds ratio of 5.00. The solid media agreed almost perfectly (κ=0.922), indicating redundancy.

Liquid media, particularly Brucella and thioglycolate broths, should be prioritised for bedside inoculation of ocular specimens. When specimen volume is limited, liquid-only inoculation avoids missing clinically significant pathogens.