Author's reply: To PMID 23202390.

Sir,

We thank the authors for taking interest in our study.[1]

The authors say that inclusion of Age-related macular degeneration (AMD) and non-AMD eyes in the same report can produce misleading results. We acknowledge this in our discussion and emphasize that the underlying etiology is an important predictor of long-term vision.

The authors raise a point regarding the natural history of visual improvement in eyes with AMD experiencing submacular hemorrhage (SMH).[23] A variability of 20-70% success, as pointed out by the authors, is extremely large. This exemplifies that visual outcomes associated with mere observation in SMH can be very unpredictable. Hence, we argue that with such a large range, most results will appear to mimic the natural history, as do results from all three arms of our study. In addition, our study cannot be compared to the natural history studies of the 1990s as polypoidal choroidal vasculopathy was not a well-defined entity then.

The authors also cite results from the Submacular Surgery Trial (SST), to discredit our results.[4] However, there are many differences between our approach and that adopted in the SST. For example, mechanical damage to photoreceptors during removal of the choroidal neovascular (CNV) membrane itself may be responsible for visual loss and may explain differences between the studies.

Citing the study by Bennett et al., the authors erroneously state that mean visual acuity improved to 20/35 in traumatic SMH.[3] Case numbers 19 and 20 had large and thick SMH due to trauma with extremely poor outcome. Berrocal et al. had myriad causes including Proliferative diabetic retinopathy (PDR), Central retinal vein occlusion (CRVO), and Presumed ocular histoplasmosis syndrome (POHS) as causes of non-AMD SMH.[2] Due to the inherent heterogeneity in the etiologies and characteristics of SMH, it is unfair to compare the non-AMD group from Berrocal et al.'s study with our study.

The authors draw attention to the complications, namely retinal detachment as a complication of vitrectomy. We would expect fewer complications in the current era with use of small gauge vitrectomy techniques.

The authors misinterpret results from Table 2 in our study. Though both follow-up period and anatomical outcome coexist in Table 2, it does not imply that it took 6.5 months for the hemorrhage to clear in all the eyes. The authors also mention that the SMH looks worse in Fig. 3b than at presentation (3a). We disagree and believe that it is more dispersed which, in our opinion, helps it reabsorb much faster.

There are many studies available to determine the utility of pneumatic displacement in AMD with SMH. The authors surprisingly choose a retrospective study with a sample size of only 15.[5] We do not have any similar groups to the ones cited by the authors. Hence, we do not understand the relevance of such a poorly powered study in the context of our results.

Though we realize that well-designed Randomised controlled trials (RCTs) are needed to determine best approaches to manage SMH, we would like to refer the readers to an excellent review article to serve as a reference till such time as an RCT is available.[6]