Purpose : Under-five mortality (U5M) rate is disproportionately high in Sub-Saharan Africa. Prior studies have revealed an association between childhood blindness and U5M. Employing CHAMPS procedures which includes postmortem minimally invasive tissue sampling and testing, we studied the feasibility for eyes to provide supportive diagnostic information for determining etiologies of U5M.
Methods : Prospective assessment of the feasibility of photography and tissue sampling of postmortem eyes for identifying disease patterns correlating with ophthalmic disease and causes of U5M. Randomly selected cases of stillbirth and under-5 death from Kenya were enrolled. External, anterior, and posterior segment images were obtained with a portable camera, graded by 2 independent ophthalmologists, and adjudicated by a third ophthalmologist. Intra-class correlation coefficients (ICC) were calculated to determine the inter-rater reliability (IRR) of the grading scheme. Conjunctival specimens were obtained for impression cytology and molecular PCR detection of pathogens.
Results : Of 71 subjects (142 eyes) that underwent ophthalmic photography and standardized grading, external photos predominantly showed periorbital edema (16%), skin mottling (14%), ecchymosis (13%), and periorbital erythema (11%). Anterior segment findings were predominantly conjunctival injection (30%), corneal opacity (23%), and conjunctival icterus (16%). Posterior segment findings were predominantly macular whitening (51%), retinal fold (34%), and optic nerve pallor (9%). The ICC for posterior segment findings was 0.68, indicating moderate IRR. The ICC for external and anterior segment findings were < 0.50, indicating poor IRR for both measures.
Of 50 subjects (100 eyes) in which impression cytology was performed, goblet cells (78%), mucin spots (24%) and squamous metaplasia (5%) were observed. Other findings were acute inflammatory cells (4%), viral inclusions (4%), bacteria morphologies (7%), and fungal elements (3%).
Conclusions : In the postmortem assessment of causes of U5M, an ophthalmic surveillance protocol was implemented and feasible with moderate IRR for posterior segment findings on photography and gradable impression cytology specimens in most eyes assessed. Further studies of these diagnostics in relation to time-of-death may provide further insight into whether ocular findings may contribute to causes of U5M.