Effect of Hyperbaric Oxygen Therapy on Central Retinal Artery Occlusion: Evidence from Observational Studies

Dear Editor,

We read with interest the systematic review and meta-analysis conducted by Rosignoli et al. [1] concerning the effects of hyperbaric oxygen therapy (HBOT) on visual outcomes following central retinal artery occlusion (CRAO). Their study included three retrospective, comparative studies [2] in the meta-analysis section and concluded no evidence of a beneficial effect of HBOT on visual acuity (VA) improvement following CRAO.

Regrettably, Rosignoli et al. [1] mixed up the VA values between the HBOT and control groups in the contributing study by Beiran et al. [2]. Specifically, the study by Beiran et al. [2] reported better final VA and greater improvement in VA in the HBOT group compared to the control group. However, Rosignoli et al. [1] presented the opposite findings. Their misinterpretation resulted in a potentially incorrect conclusion of nonsignificance. Additionally, the study by Beiran et al. [2] encompassed all types of retinal artery occlusion, including cases of branch retinal artery occlusion (20%). This deviation from the exclusive focus on CRAO was inconsistent with the primary focus of the systematic review and meta-analysis [1].

Here, we reconducted the meta-analysis for the two outcomes, final VA and improvement in VA, based on the data from Rosignoli et al. [1], while ensuring the correct placement of the VA values from the study by Beiran et al. [2] into their respective HBOT and control groups. We pooled the study-specific estimates using inverse-variance weighting with random effects via the Cochrane RevMan Web ver. 6.2.0 (https://revman.cochrane.org/).

Our meta-analysis demonstrated that the final VA was better in the HBOT group than the control group (Fig. 1A). Likewise, the improvement in VA was greater in the HBOT group than the control group (Fig. 1B).

Fig. 1

Meta-analysis of the effects of hyperbaric oxygen therapy on (A) final visual acuity and (B) improvement in visual acuity following “central” retinal artery occlusion. The contributing study by Beiran et al. [2] included cases of branch retinal artery occlusion. Visual acuities are expressed in logarithm of the minimum angle of resolution. SD = standard deviation; IV = inverse variance; Random = random-effects model; CI = confidence interval.

By correctly placing the VA values into their respective comparison groups, our revised meta-analysis suggests that HBOT might have a beneficial effect on VA improvement following CRAO. This contrasts with the results reported by Rosignoli et al. [1], whose nonsignificant results were flawed and have unfortunately been cited by a few others [3–5], incorrectly asserting that HBOT offers no benefit for CRAO. Despite this, our meta-analysis was still limited by the use of data from observational studies, which are open to confounding and selection bias, and the inclusion of non-CRAO cases. Thus far, no relevant randomized controlled trial has been conducted. Therefore, our results should be cautiously interpreted. Large-scale randomized controlled trials are warranted to establish the effectiveness of HBOT as a treatment for CRAO.

In conclusion, we respectfully contend that the meta-analysis conducted by Rosignoli et al. [1] does not accurately reflect the effect of HBOT on VA improvement following CRAO.