Detection of anomalies in the red reflex test requires adequate training

Midwives should receive appropriate training to enable them to reliably detect an abnormality in the red reflex, so as to detect potentially sight- (and possibly life-) threatening visual anomalies in newborns.

Midwives who had completed the relevant Newborn and Infant Physical Examination (NIPE) training in the UK, were tested to determine whether they could reliably detect an abnormality in the red reflex of a model eye. Their results were compared to a group of undergraduate optometry students in their final year, who had considerably more experience with ophthalmoscopy.

Two groups of adult participants, 27 post-graduate midwives and 10 undergraduate optometry students with different levels of experience in ophthalmoscopy were recruited. Participants were required to perform the red reflex test repeatedly on a set of 10 model eyes with interchangeable simulated retinas, that produced both normal and abnormal 'red' reflexes, under controlled conditions.

A total of 1,810 observations were recorded across both groups of participants. Both the sensitivity and the specificity for distinguishing between normal and abnormal reflexes were higher for the optometry students (sensitivity 77.5 per cent and specificity 96.7 per cent) than the midwives (sensitivity 53.9 per cent and specificity 75.4 per cent). The positive predictive values were 21.7 per cent and 67.4 per cent for the midwives and optometry students, respectively. The negative predictive values were 92.8 per cent for the midwives and 98.0 per cent for the optometry students.

Despite completion of required training on the red reflex test, optometry students outperformed midwives in the detection of anomalous red reflex test simulations from a model eye. This result is likely due to differences in training in ophthalmoscopy between the two groups of participants. Additional training in the conduct of the red reflex test for midwives is therefore recommended.