Impact of patient face mask use on endophthalmitis after intravitreal anti-VEGF injections

To compare the incidence of endophthalmitis after intravitreal anti-vascular endothelial growth factor (anti-VEGF) injections with and without patient face masking. Two-centre retrospective cohort study Patients receiving a total of 28 501 injections (period 1 before masking: n = 13 863; period 2 after masking: n = 14 638) Periods before and after implementation of patient masking were compared: period 1 (July–December 2019) and period 2 (July–December 2020). All patients requiring treatment for endophthalmitis following intravitreal anti-VEGF injections were reviewed. Endophthalmitis risks were compared. Statistical simulations were run to determine the number of injections or endophthalmitis cases required to detect a statistically significant difference between both periods. Five patients (0.036%) had endophthalmitis in period 1 compared with 7 patients (0.048%) in period 2. Odds ratio, 95% confidence interval, for risk of patient masking was 1.326, 0.421–4.179 (p = 0.63). Three patients (0.022%) in period 1 had culture-positive endophthalmitis compared with 2 patients (0.014%) in period 2. Risk of patient masking for culture-positive endophthalmitis was 0.631, 0.105–3.779 (p = 0.61). Assuming similar rates of endophthalmitis, 476,806 injections in both groups would be required to detect a significant difference. Alternatively, assuming that masking would increase endophthalmitis risk in period 2, a significant difference would require twice as many cases (n = 15, 0.102%) for a risk of 2.843, 1.033–7.825 (p = 0.043). The implementation of patient masking in clinical practice did not significantly alter the rate of endophthalmitis following intravitreal anti-VEGF injections. Comparer l'incidence d'endophtalmie au décours d'injections intravitréennes d'anti-VEGF (facteur de croissance endothélial vasculaire) avec ou sans port du masque par le patient. Étude de cohorte rétrospective réalisée dans 2 centres. Patients qui ont reçu un total de 28 501 injections (période 1 : sans masque; n = 13 863; période 2 : avec masque; n = 14 638). Ont été comparées les périodes avant et après l'implantation du port du masque par les patients: période 1 (juillet–décembre 2019) et période 2 (juillet–décembre 2020). On a passé en revue les dossiers médicaux de tous les patients qui ont dû être traités en raison d'une endophtalmie après avoir reçu des injections intravitréennes d'anti-VEGF. Les risques d'endophtalmie ont alors été comparés. On a effectué des simulations stat