**6. Diagnostic investigations**

Sampling of the intraocular fluid for Gram stain and culture is recommended before treatment is initiated. However, empiric treatment should begin promptly while the specimen is being analysed. Working closely with the microbiology department is mandatory in managing these cases.

Intraocular fluid samples should be sent for Gram and Giemsa staining, culture, and sensitivity. Gram stain is positive in approximately 45% of endophthalmitis cases [59]. Vitreous sampling has a higher positive culture rate than aqueous humour, with vitrectomy samples being more culture positive than vitreous tap [60].

Polymerase chain reaction (PCR) is useful in culture negative cases. It is effective in detecting fastidious bacteria, provides a result in a few minutes and is more sensitive than cultures [61, 62]. However multiple samples have to be sent for each micro-organism that the clinician suspects. Biome representational in silico karyotyping (BRiSK) is a new technique for detection of any DNA-based life form and can

detect micro-organisms in PCR negative samples [63]. However, its role in BRI is not established.

Although the majority of glaucoma specialists recommend conjunctival sampling, it must be appreciated that conjunctival culture results in BRI do not generally match the culture results from anterior chamber and vitreous humour samples [32, 39, 64]. Care must be taken not to tear the conjunctiva over the bleb area when taking samples.
