# Interventions for submacular haemorrhage: A systematic review and network meta‐analysis of controversies—On behalf of the Spanish Vitreo‐Retinal Society (SERV)

## TL;DR

This study reviews treatments for submacular haemorrhage, finding that non-vitrectomy methods may offer better visual outcomes with fewer complications.

## Contribution

The study provides a network meta-analysis of SMH interventions, highlighting controversies and the lack of a standard treatment.

## Key findings

- Non-vitrectomy therapies showed better visual acuity outcomes and fewer complications.
- Observation had the highest probability of being most effective for BCVA, but with limited data.
- Vitrectomy-based treatments achieved better anatomical results despite higher risks.

## Abstract

This systematic review aims to evaluate and synthesize the existing literature on the interventions used for submacular haemorrhage (SMH), highlighting the controversies and differences in clinical practice.

A systematic review was conducted following the PRISMA guidelines. A comprehensive search was performed across multiple databases, including MEDLINE, EMBASE and Cochrane Library, to identify studies on SMH treatment. Inclusion criteria encompassed randomized controlled trials, cohort studies and case series that focused on different therapeutic interventions. Data on functional outcomes, efficacy and safety of the interventions were extracted and analysed.

The review included 150 studies, of which 38 were included in the network meta‐analysis. The analysis of best corrected visual acuity (BCVA) Included 26 studies, 20 interventions and 2125 eyes. Heterogeneity was moderate (
I
2
 = 28.9%). Non‐vitrectomy therapies showed better BCVA outcomes and fewer complications (e.g. retinal detachment, vitreous haemorrhage), while vitrectomy‐based treatments achieved better anatomical results. According to P‐score ranking, “Observation” had the highest probability of being most effective for BCVA (P‐score = 0.8051), followed by anti‐VEGF monotherapy and non‐vitrectomy combinations. However, this result should be interpreted cautiously, as the “Observation” group was based on only two studies (26 eyes) with clinical heterogeneity. No publication bias was detected (Egger's test p = 0.582).

There is no consensus on a standard evidence‐based treatment for SMH. Minimally invasive strategies are promising, but factors such as timing, lesion size and anti‐VEGF use remain critical. Further large‐scale randomised trials are needed to define optimal management.