Design: phase 3, multicentre, double-masked, randomised, placebo-controlled trial
Population: 145 subjects with dim light disturbances, mesopic pupil 5 mm or more and mesopic low-contrast acuity 30 ETDRS letters or worse (25 post-keratorefractive)
Primary outcome: improvement in mesopic low-contrast best-corrected distance visual acuity
Effect: responders 21% versus 3% on placebo at day 15 (P<0.01); 13% versus 3% at day 8 (P<0.05)

One hundred and forty-five people with self-reported dim light disturbances, a mesopic pupil diameter of 5 mm or more, and mesopic low-contrast best-corrected acuity of 30 ETDRS letters or worse were randomised 1:1 in a phase 3, multicentre, double-masked trial to 0.75% phentolamine ophthalmic solution or placebo, one drop in each eye nightly for 14 days. Twenty-five were post-keratorefractive.

Mesopic low-contrast acuity improved in 13% of treated subjects against 3% on placebo at day 8 (P<0.05), and 21% against 3% at day 15 (P<0.01). Reported severity of dim light disturbance fell, as did glare, halos and starburst, at both time points (P<0.01). Conjunctival hyperaemia and vital signs did not differ between arms. A post-hoc analysis of the post-keratorefractive subgroup showed the same direction on both acuity and patient-reported outcomes.

The mechanism is the appeal: phentolamine is an alpha-adrenergic antagonist, and reducing mesopic pupil diameter cuts the peripheral corneal rays that produce halo and starburst. That is a rational answer to a complaint for which the honest current answer is usually that nothing can be done — a patient several years after refractive surgery who cannot drive at night.

The limits are substantial and worth being straight about. Fourteen days of treatment tells you nothing about a symptom people live with for years, the acuity endpoint is a surrogate for the driving complaint, and the responder rate — one in five — means four in five did not improve on the primary measure. The authors' own conclusion is that further investigation is warranted, not that the problem is solved.

Keep this in mind for the post-refractive patient with night-driving glare, but do not promise a treatment yet
Note the entry criteria: mesopic pupil 5 mm or more, and measurably reduced mesopic low-contrast acuity — not all glare qualifies
Check the patient's refraction, tear film and any posterior capsule opacification first; those are correctable causes of the same complaint
Ask specifically about night driving, which is the function the symptom actually costs
Watch for longer trials before counselling about sustained benefit; this was 14 days

Don't overread it

Fourteen days of treatment and a surrogate acuity endpoint — this does not show sustained relief of night-vision symptoms.

The statistics, in plain English

A responder rate of 21% against 3% is a real separation from placebo, but it means roughly four in five treated subjects did not meet the acuity threshold. The patient-reported improvements are more clinically meaningful and also more susceptible to expectation effects, even in a masked trial, because a drop that constricts the pupil has a perceptible effect. The post-keratorefractive analysis was post-hoc in 25 people, so it supports the hypothesis rather than establishing benefit in that group.