Statistics from Altmetric.com
It has recently been reported that there has been a trend in the United Kingdom towards fewer children requiring strabismus surgery.1 Unless there has been a change in the incidence of childhood strabismus, it may reflect a change in the clinical practices of ophthalmologists, optometrists, and orthoptists. One such practice is the correction of associated refractive errors. The correction of a hypermetropic refractive error is an important aspect in the management of esotropia.1–4 Although the full hypermetropic correction is usually prescribed for children with an esotropia, it is unclear whether this is, or has been, uniform in practice. Indeed, certain texts suggest a reduction in the hypermetropic correction following a cycloplegic refraction, to take into account so called accommodative tonus.4 Because of the effect of an undercorrection of a hypermetropic refractive error on the angle of esotropia, we were interested to determine whether a change in spectacle prescribing practice had had any effect on the number of patients undergoing squint surgery.
Data were collected from the computerised hospital attendance statistics and operations at the Royal Liverpool Children’s NHS Trust Hospital, which were available from the financial year 1988–9 onwards. The annual number of ophthalmic referrals and operations, were collected within two time periods: a 4 year period from 1988–9 to 1991–2 (period 1) and a 7 year period from 1994–5 to 2000–1 (period 2). Before 1992, it had been the practice to prescribe spectacles with a 0.5–1 dioptre reduction in the spherical component of the cycloplegic refractive error. From the period 1994 onwards, it was the practice of the consultant for the full cycloplegic hypermetropic error to be prescribed. Data were not collected between 1992 and 1994 to allow for a possible washout period for those patients who had previously been prescribed an undercorrected hypermetropic correction and were undergoing surgery between 1992 and 1994. The total number of eye and squint operations performed per 100 patient referrals, and the ratio between the squint and non-squint operations performed were calculated. The difference between the two periods was compared using the Fisher’s exact test and confidence intervals calculated by the Miettinen approximation method.
During periods 1 and 2, 4514 and 10 314 new patients were seen, of whom 23.9% and 23.6% underwent some form of ophthalmic surgery (p >0.5). There was a significant difference between the mean number of squint operations per 100 patients in period 1 compared to period 2 (14.8% and 9.3%, p <0.0001, proportion difference 0.060, 95% CI 0.047 to 0.073). There was also a significant difference in the mean number of squint operations per 100 total operations performed in period 1 compared to period 2 (61.9% and 38.2%, p <0.0001, proportion difference 0.243, 95% CI 0.203 to 0.283).
Although it was not possible to examine the population profiles in the time periods studied, there had been no known change in the population profile or referring practices. The hospital data retrieval system, unfortunately, did not permit us to differentiate between a new patient seen with a squint and another seen with a non-squint problem. While the proportion of children who required surgery had remained constant, there had been, similar to a recent report,1 a reduction in the number of squint operations being performed, with a corresponding increase in the proportion of non-squint operations. We agree that there were likely to have been a variety of other factors contributing to the reduction in strabismus surgery. However, this trend may reflect in part, a change in practice whereby children with strabismus who had a hypermetropic refractive error were, during the latter period, prescribed their full hypermetropic correction.