PurposeTo assess the safety and efficacy of an eye drop combining osmoprotectants, carboxymethylcellulose and hyaluronic acid (O/CMC/HA) in reducing symptomatic, moderate to severe dry eye, compared with HA.MethodsIn this investigator-masked, randomised study, patients instilled 1-2 drops/eye of O/CMC/HA or HA (2-6 times/day) for 3 months. Primary endpoint: mean change in Global Ocular Staining Score (GOSS) from baseline at day 35. Noninferiority of O/CMC/HA was tested in the per-protocol population; if achieved, superiority was tested in the intent-to-treat population. Secondary efficacy endpoints: mean change from baseline in GOSS, Ocular Surface Disease Index (OSDI), Schirmer score, tear break-up time (TBUT), corneal/conjunctival staining, conjunctival hyperaemia, symptoms, and patient/investigator assessments.ResultsBaseline characteristics were comparable between groups (n=40 each). O/CMC/HA was noninferior (and not superior) to HA based on similar GOSS reductions from baseline at day 35 and month 3 in both groups (P=0.778, day 35, per-protocol population). Overall, O/CMC/HA and HA provided similar reductions in OSDI, Schirmer score, TBUT, corneal staining and hyperaemia from baseline at 35 days (P≥0.155). More patients reported less severe stinging/burning, sandiness/grittiness, and painful/sore eyes at month 3 with O/CMC/HA (P≤0.039), and more rated the dropper bottle easy to use (87.5%), compared with HA (46.2%; P=0.002). Other patient and investigator assessments were similar between groups. O/CMC/HA and HA were well tolerated.ConclusionsO/CMC/HA is noninferior to HA in improving objective signs of dry eye, with potential advantages for subjective symptoms and patient acceptance.