Background There is an increasing demand for routine assessment of surgical outcomes. However, for assessments to be valid, it is essential to understand the influence of patient characteristics on outcomes. The VF-14 visual function instrument offers a possible means of monitoring outcomes of cataract surgery. This article examines lessons learned in the course of an attempt to do this.

Methods One thousand and ninety-eight patients (41 per cent male and 59 per cent female) undergoing cataract extraction in 29 private hospitals within the British United Provident Association (BUPA) chain in the United Kingdom completed the VF-14 before and at 4 months following surgery. Outcome was measured in terms of absolute post-operative score achieved. The demographic factors associated with post-operative VF-14 score were examined using multivariate analysis, adjusting for baseline score.

Results Women undergoing cataract extraction had worse pre-operative visual function than men. Post-operative VF‐14 score was stable as the age of the patient increased, until age 85, after which it declined. The reduction in final outcome in older people was explained in part by their worse post-operative scores but also was associated with increasing age independent of pre-operative score.

Conclusion First, although many patients undergoing cataract extraction at an advanced age will achieve excellent results, older age does reduce the outcome that can be expected. Consequently, comparisons of performance of units or surgeons undertaking cataract surgery must take account of the age spectrum being treated. It is not enough simply to add a health status measure to an existing administrative database. Further research is needed to understand the reasons for the poorer outcomes at older ages, in particular the role of co-existing eye disease and development of data systems must take account of the additional information required to make appropriate adjustments.