Bill Harvey: Live and direct

A good clinical trial might reveal a statistically significant positive outcome from an intervention in one target group and no such outcome in a control target group, where both groups are matched and randomly allocated.

I am always going on about what sometimes seems the universal prescribing of ocular lubricants in non-symptomatic patients. Some of you have contacted to say that their use is warranted in that the drops will be prophylactic and prevent the onset of future symptoms. The evidence base here is sparse. Also, the ‘20-20-20 rule’ for computer users seems sensible, but where is the evidence base?

All well and good, but examples are rife in eye care where proof of a negative impact might be useful. Take, for example, the long-held view that there is no published evidence that ‘a lazy eye’ has a negative impact on normal development, a statement used to argue against a comprehensive paediatric eye health screening service. Proof of such an impact would require a group of children to be deliberately blurred in one eye and followed for years. Look at longitudinal studies in myopia management. There comes a time where the benefit of intervention in one group makes the non-intervention in the control group questionable.