No abstract

grades: one by a consultant, three by associate specialists, one by a registrar and two by senior house officers. Details relating to the method of peribulbar injection have been obtained for six cases. A sharp 25 mm 25G needle was employed in four cases and a 38 mm 25G retrobulbar needle for two. All the anaesthetists used two injections sites, namely the infero-temporal or inferior approach followed by a second injection via the medial, supero-nasal or superior approach. Four of the perforations occurred in either the superior or supero-nasal aspect of the globe, one occurred nasally and in one case the site of the perforation was not clear.The possibility of a perforation was entertained before or during surgery for three cases: one was noted to have a hyphaema before surgery, one a very soft eye before surgery and in one no red reflex was present after expression of the lens. The diagnosis of a perforation was made within the first week for six of the seven patients, all of whom had a vitreous haemorrhage. Four had either cryotherapy or laser to close a retinal break without sub-retinal fluid, two required pars plana vitrectomy with fluid-gas exchange and endolaser, while one patient refused further surgery. The final acuity has been reduced to perception of light for two patients.Our experience supports the belief that the incidence of ocular perforation during peri bulbar anaesthesia is rising and the figure is certainly higher than the 0.006% incidence quoted by Davis and Mande1.2 Although most peribulbar injections are performed by consultant anaesthetists in this area, most of the perforations were caused by other grades. This implies that the training and supervision of those performing peribulbar blocks could be improved, and a case could be made for avoiding the superior or supero-nasal routes. Fresh vitreous haemorrhage on the first post-operative day is a hallmark of ocular perforation. With early recogni tion that a perforation has occurred the visual prognosis is better, although the final visual outcome is mixed.

Need for more cautious analysis EDITOR,-D A H Laidlaw and colleagues describe the effect of the introduction of the sight test fee on the number of referrals to Bristol Eye Hospital.' They use the numbers of referrals for 1984-8 to predict the expected numbers of referrals in 1989-92. This use of their data would have been appropriate if the introduction of the fee on 1 April 1989 had been made without prior warning. The government's proposal for the sight test fee had, however, been made public on 25 November 1987.2 This announcement (and subsequent confirmations) would be expected to have had two effects. Firstly, the high figure for referrals in 1988 may partly simply be due to the increased general publicity about sight tests. Secondly, some people who would have had an eye test after 1 April 1989 may have decided to have one beforehand instead. These "early" sight tests would have increased the figure for referrals in 1988 and reduced that in 1989. The authors consider whether early referrals received in 1988 could account for a reduction in referrals from 1989 onwards. They reject this idea on the grounds that the number of referrals in 1988 was a mere 343 (5 6%) above that expected. Ifthese 343 are regarded as early referrals and are taken from the total for 1988 and added to the total for 1989, however, then the fall in referrals from 1988 to 1989 is not by 15-1% but by only 4-8%. Whether this transfer of numbers of referrals is valid is clearly arguable, but this manoeuvre illustrates that a more cautious analysis might have been appropriate in the study. In particular, it would have been preferable not to have used data for 1988 (the year before introduction of the fee) to predict the numbers of referrals afterwards. If the figure for 1988 is omitted from the analysis the actual numbers of referrals for 1989-92 all fall within the recalculated 95% confidence intervals for the predicted numbers. This implies that the authors have not convincingly shown that the numbers of referrals have been less than expected since the introduction of the sight test fee.