Background: In general practice, infectious conjunctivitis is a common and mostly (64%) self-limiting disorder. In case of an aberrant course or severe symptoms, a general practitioner may take a culture. Direct inoculation is considered the reference standard, but usually a swab is sent to a laboratory.

Objectives: To compare the diagnostic performance of the swab, transported by surface mail with direct inoculation.

Methods: 19 general practitioners took two samples of the conjunctiva from 88 patients with symptoms suggestive of infectious conjunctivitis by rolling a cotton swab across the conjunctiva of the lower fornix. One swab was used to inoculate three agar plates directly, while the other was sent in a Stuart medium to the laboratory and inoculated at the time of arrival. The numbers of positive cultures of both methods were compared.

Results: A pathogen was found in 31 of 88 samples (35% (95% CI 26 to 46)). Surprisingly, the number of positive cultures was higher for the Stuart medium (27/88) than for direct inoculation (23/88). The difference was 4.5% (90% CI 0 to 12, p = 0.388; one-sided McNemar test for paired proportions). In five of the 19 samples that were positive in both tests, the cultured pathogens were different.

Conclusions: The Stuart medium detected more bacteria than direct inoculation. The lower 90% CI, testing non-inferiority at p = 0.05, indicates that it is unlikely that the Stuart medium misses any positive cultures compared with direct inoculation.