Statistics from Altmetric.com
Editor,—We read Tanner et al's paper on the predictive value of vitreous pigment (Schaffer's sign) for retinal breaks in posterior vitreous detachment1 with great interest. Based on their figures, patients who have a negative Schaffer's sign had a 1% chance of having a retinal tear or hole and a 0.5% chance of having a lesion where prophylaxis was thought to be appropriate. Thus, Schaffer's sign has a negative predictive value of 99% in their series. They go on to recommend that if vitreous pigment is present then the patient should be referred for urgent vitreoretinal opinion while those with no pigment should be referred on a less urgent basis.
We would like to put these findings in perspective. The incidence of retinal breaks in patients aged 10 years or more who do not have any history of ocular disease is 6–14%.2 Retinal breaks have been found in 37/250 (14.8%) of necropsy eyes with posterior vitreous detachment by Foos.3 The incidence of retinal detachment is approximately 12/100 000 of the general population per year.4 This suggests that less than 0.2% people with a retinal break eventually have a detachment of the retina. This value may be higher in patients with a symptomatic posterior vitreous detachment; however, it is reasonable to conclude that only a minority of retinal breaks will go on to cause a retinal detachment. Prophylactic treatment of retinal breaks by laser or cryotherapy is not without complications; also detachments can occur in eyes that have had prophylactic treatment.5 Byer has reported that retinal breaks in unoperated eyes with posterior vitreous detachment that need treatment can be followed up without treatment, with only a minority progressing to retinal detachments.6
We have a test that has a negative predictive value of 99%. We know that only a minority of patients who have a retinal tear or hole actually benefit from prophylactic treatment. Can we still justify referring all patients with a posterior vitreous detachment and no vitreous pigment for a specialist examination or even a follow up examination in the light of this knowledge?
The appropriate recommendation would be that all patients presenting with posterior vitreous detachment, no vitreous pigment, and no retinal tears or holes at initial examination can be safely discharged with an explanation of the warning symptoms which should prompt the patient to reattend.
Editor,—We thank Gupta and Prasad for their interest and comments on our recent paper. We agree that the majority of retinal breaks probably do not progress to cause retinal detachment but suggest caution in the interpretation of data relating to asymptomatic rather than symptomatic tears, the latter having been shown to be associated with subsequent retinal detachment in approximately 30% of cases. It is the management of recent onset, symptomatic posterior vitreous detachment (PVD) and associated retinal breaks which we addressed in our study. The rationale for treating fresh, symptomatic retinal breaks has been reviewed by numerous authors and is best summarised in the recent preferred practice pattern document produced by the American Academy of Ophthalmology.
The purpose of our study was to provide help to those practitioners seeing large numbers of patients with acute PVD, but who do not feel confident in the use of indentation ophthalmoscopy. The presence of vitreous pigment in patients presenting with acute PVD is indeed highly predictive of the presence of a retinal break, but a thorough retinal examination is still necessary.
In our series only one patient represented with a retinal break which had not been identified during initial indentation ophthalmoscopy. We agree that no routine follow up examination is required in most cases, provided a retinal break has been confidently excluded with indentation ophthalmoscopy and the PVD is judged to be complete, but that patients should be warned to reattend if further symptoms occur.