Six publications, referring to 3 RCTs (n=4,194), met the inclusion criteria. The authors also referred to 4 and 5 studies though, presumably, they meant publications.
Most of the studies were categorised as A or B for methodological quality. All of the studies checked medication compliance. The number of study participants ranged from 8 to 3,711.
Aspirin alone neither prevented the development of high-risk proliferative retinopathy (relative risk, RR=0.97, 95% confidence interval, CI: 0.85, 1.11), nor increased the risk of vitreous haemorrhage (RR 1.0, 95% CI: 0.8, 1.3) (1 large study).
The authors stated that there was no significant difference in the severity of vitreous or pre-retinal haemorrhage for aspirin versus placebo or aspirin-dipyridamole versus placebo (1 study; data not reported). There was no statistically significant difference in the 5-year vitrectomy rates in the aspirin group compared with the placebo group (5.4% and 5.2%, respectively), although the authors reported that the CI (which was not reported) includes important benefit and harm. There was no significant difference in the aspirin alone group and the aspirin-dipyridamole group in the development of microaneurysms. The mean annual increase in microaneurysms was significantly higher (P=0.02) in the treated group than the placebo group.
In a two-period crossover study of 8 patients there was a mean aspirin-placebo treatment difference of 21% (95% CI: 4, 38, P=0.03).